BICSI Ict-Today

ICT TODAY
THE OFFICIAL TRADE JOURNAL OF BICSI

October/November/December 2020
Volume 41, Number 4
A high-speed wireless highway for tomorrow’s

fastest devices and services.
PLUS:
+ Li-Fi: The Future of LED Lighting Technology
+ Leveraging Infrastructure to Optimize the Edge
+ Multimode Fiber Technology, Latest Market
Trends and Standards
Your fusion splicer.
Our connector.
New Corning FuseLite® 2 Splice-On
To learn more, visit
Connectors bring it all together. corning.com/fuselite2
or contact your Corning
This is the high-performance matchup you’ve been waiting for. Our new FuseLite 2
Splice-On Connectors let you combine the reliable, high-quality performance you
representative today.
trust from Corning with the flexibility and familiarity of your favorite fusion splicer.
These easy-to-assemble connectors are quick to install and come packed with
features that make them a great choice for any new installation, repairs and
restorations, even reflectant-sensitive applications. They’re available for single-
fiber termination in SC, LC, and ST® formats for both single-mode and multimode.
© 2020 Corning Optical Communications. All rights reserved. LAN-2728-A4-BEN / July 2020
OCTOBER/NOVEMBER/DECEMBER 2020
Volume 41, Issue 4 contents
FROM THE PRESIDENT
05 Essential ICT Professionals 36 Protecting the Future at the Edge: Addressing

and Emerging Technologies the Risks of Emerging Edge Data Centers: Seventy-
By Todd W. Taylor, RCDD, NTS, OSP five percent of enterprise-generated data is
expected to be created and processed outside
COVER ARTICLE of traditional data centers. Discover the importance
06 Wi-Fi 6E for the 10 Gb Era: Wi-Fi 6E is described of physical protection via critical practices for ensur-
ing that EDCs remain operational, secure, and
as a private highway for the latest and greatest
reliable. Without critical practices in place, the
devices and services. Discover the unique advan-
emerging technologies EDCs aim to support will
tages of this technology, how it benefits from 6 GHz
struggle needlessly as they gain adoption.
unlicensed spectrum, how it impacts PoE, why the
By Jacques Fluet
lines between it and 5G are blurring, and
46 Multimode Fiber Technology, Latest Market

much more.
By Sundar Sankaran, Ph.D.
Trends and Standards: This article is designed in two
16 Li-Fi: The Future of LED Lighting Technology: parts to provide the most current and comprehen-
sive knowledge about MMF. In part one, the trends
Imagine making an internet connection as simple
behind MMF’s success story are explored. Part two
as turning on a light switch. The possibilities of how
provides the most recent standards update, provid-
data can be used are exciting and endless. Learn
ing ICT professionals the capability to offer their
everything about Li-Fi in this in-depth study, including
customers the best this technology has to offer
how it works, applicable standards for installation,
now and into the future.
cabling infrastructure solutions, and current
By G. Mabud Choudhury and John Kamino, RCDD
challenges for mass deployment.
66 The Impact of ICT on Healthcare Simulation:

By Miro Sino, RCDD, P.Eng, LC, LEED AP
26 Leveraging Infrastructure to Optimize the As health care simulation technology is increasingly

making its way into the delivery of medical care
Edge: Explore the ways to leverage infrastructure
and medical educational institutions, its future
that impacts the deployment, monitoring, manage-
depends on reliable networks designed, installed,
ment, and operation of an edge data center. Infra-
and supported by ICT professionals. Learn about
structure choices address common challenges
installation methods, cabling, best practices, and
related to security, environmental conditions, and
much more.
reliability, thereby ensuring uptime for the network
By Patti Fisher, RCDD, DCDC, OSP, RTPM, WD
and those systems dependent on the network.
By Brian L. Kelly and Mark Dehmlow
SUBMISSION POLICY
ICT TODAY is published quarterly in January/February/March, April/May/June, July/August/September, and October/November/December by BICSI, Inc., and is sent to BICSI members, RCDDs, RTPMs,
DCDCs, BICSI Installers and Technicians and ESS, NTS, OSP, RITP and WD credential holders. ICT TODAY subscription is included in BICSI members’ annual dues and is available to others through a purchased
yearly subscription.
ICT TODAY welcomes and encourages submissions and suggestions from its readers. Articles of a technical, vendor-neutral nature are gladly accepted for publication with approval from the Editorial Review Board.
However, BICSI, Inc., reserves the right to edit and alter such material for space or other considerations and to publish or otherwise use such material. The articles, opinions and ideas expressed herein are the sole
responsibility of the contributing authors and do not necessarily reflect the opinion of BICSI, its members or its staff. BICSI is not liable in any way, manner or form for the articles, opinions and ideas, and readers are
urged to exercise professional caution in undertaking any of the recommendations or suggestions made by authors. No part of this publication may be reproduced in any form or by any means, electronic or mechanical,
without permission from BICSI, Inc.
ADVERTISING: Advertising rates and information are provided upon request. Contact the BICSI Sales Department for information at +1 813.979.1991 or 800.242.7405 (U.S. and Canada toll-free)
or cnalls@bicsi.org. Publication of advertising should not be deemed as endorsement by BICSI, Inc. BICSI reserves the right in its sole and absolute discretion to reject any advertisement at any time by any party.
POSTMASTER: Send change of address notices to BICSI, Customer Care, 8610 Hidden River Pkwy, Tampa, FL 33637-1000; Phone: +1 813.979.1991 or 800.242.7405 (U.S. and Canada toll-free).
© Copyright BICSI, 2020. All rights reserved. BICSI and RCDD are registered trademarks of BICSI, Inc.
October/November/December 2020 I 3
THE OFFICIAL TRADE JOURNAL OF BICSI
BICSI BOARD OF DIRECTORS ADVERTISER’S INDEX
AEM............................................................ 77
President Todd W. Taylor, RCDD, OSP, NTS
AEM-Test.com
President-Elect Carol Everett Oliver, RCDD, DCDC, ESS America Ilsintech........... Inside Back Cover
Secretary Honorico “Rick” Ciordia, RCDD, DCDC, RTPM, PE americailsintech.com
Corning........................... Inside Front Cover
Treasurer David M. Richards, RCDD, OSP, NTS, TECH, CT corning.com/fuselite2
Global Region Director Betty Bezos, RCDD, DCDC, OSP, ESS, NTS, PMP, PE Extron......................................... Back Cover
extronclassroom.com
EMEA Region Director Barry Shambrook, RCDD
FIS............................................................... 21
Canadian Region Director Fernando Neto, RCDD fiberinstrumentsales.com
U.S. North-Central Region Director Randal Reusser, RCDD, TECH, CT Front Row................................................... 43
gofrontrow.com
U.S. Northeast Region Director William Foy, RCDD, OSP, ESS, NTS, WD
ICC............................................................. 45
U.S. South-Central Region Director Mark Reynolds, RCDD icc.com
U.S. Southeast Region Director Lee Renfroe, RCDD, ESS, TECH MaxCell...................................................... 53
maxcell.us
U.S. Western Region Director Pat McMurray, RCDD, DCDC, NTS, OSP, PMP
nVent Caddy............................................ 41
Executive Director & Chief Executive Officer John H. Daniels, CNM, FACHE, go.nvent.com/datacom
FHIMSS
EDITORIAL REVIEW BOARD BICSI INFORMATION

EMEA Conference.................................... 65
Chris Scharrer, RCDD, NTS, OSP, CTS-D
Fall Conference........................................ 11
Jonathan L. Jew
RCDD Program.......................................... 34
F. Patrick Mahoney, RCDD, CDT Standards Program................................... 72
TDMM......................................................... 58
PUBLISHER Virtual Classroom Training........................ 78
BICSI, Inc., 8610 Hidden River Pkwy., Tampa, FL 33637-1000
Phone: +1 813.979.1991 Web: bicsi.org ICT TODAY NEEDS WRITERS

ICT Today is BICSI’s premier publication for
authoritative, vendor-neutral coverage and
EDITOR insight on next generation and emerging
technologies, standards, trends, and
Alexandra Manning, icttodayeditor@bicsi.org
applications in the global ICT community.
Consider sharing your industry knowledge
PUBLICATION STAFF and expertise by becoming a contributing
John Bechtold, Publisher writer to this informative publication.
Jamie Finne, Creative Contact icttodayeditor@bicsi.org if you

are interested in submitting an article.
Clarke Hammersley, Director, Technical Editing
Jeff Giarrizzo, Senior Technical Editor ADVERTISING SALES

Allen Dean, Technical Editor +1 813.979.1991 or cnalls@bicsi.org
4 I ICT TODAY
From the President, Todd W. Taylor, RCDD, OSP, NTS
ESSENTIAL ICT
PROFESSIONALS AND
EMERGING TECHNOLOGIES
In my previous ICT Today message, I had addressed how which is still in the early stages of development and imple-
the ICT industry is classified as an essential business during mentation, is covered in ANSI/BICSI 007-2020, Information
the COVID-19 crisis and how technicians, installers, and Communication Technology Design and Implementa-
other ICT professionals have kept emergency services and tion Practices for Intelligent Buildings and Premises. Like-
other essential industries up and running. The badge wise, all BICSI standards manuals strive to include the
of honor for these courageous frontline professionals latest and greatest technology information so that essen-
should be worn not only during times of crisis but every tial ICT professionals can be on top of their game and
day. In our industry, there are always daily network familiar with other technologies that may impact their
emergencies, spanning all vertical industries, that they mission-critical work.
must handle quickly and efficiently. These ICT professionals,
therefore, are essential workers year-round. Besides ICT Today and updated BICSI standards manu-
als, there are many other ways BICSI provides the most
Admittedly, this gives me a sense of pride serving this recent and accurate technology information including:
essential industry and its essential workers. As technolo- • B
ICSI Connect online courses
gies are advancing, changing, and emerging at a histori-
• Webinars
cally fast pace, it is also a realization of our crucial role
in further educating BICSI credentialed technicians, install- • Specialty publications on a wide range of topics
ers, Registered Communications Distribution Designers
• P
ocket-sized installation field guides covering
(RCDDs), Data Center Design Consultants (DCDCs), OSP
everything needing to be known for optical fiber
Designers, and Registered Telecommunications Project
and copper cabling and connectivity
Managers (RTPMs) about these new and emerging tech-
nologies. Consider the Wi-Fi 6E and Li-Fi articles in this very • B
ICSI’s state-of-the-art training facility
publication. It also means having to add and remember in Tampa, Florida
new acronyms to a continually growing and seemingly
Then, there are the BICSI conferences and exhibitions
endless list.
worldwide. Whether live or virtual—such as this year’s
Because networks are converging, some of the emerg- BICSI Fall Conference & Exhibition virtual online event,
ing technologies BICSI addresses span beyond traditional Leading the Future for the Connected World—emerging
ICT applications, thereby educating credentialed profes- technologies are highlighted through compelling presen-
sionals on the changing role of IT and providing insights tation sessions and demonstrations exhibited by innova-
into the work of electrical engineers/installers that can tors and users in their common quest to advance the
impact mission-critical ICT networks, troubleshooting, future of communications technologies and the
and the onsite jobs of our essential workers. ICT industry.
This further education is accomplished by ICT Today pub- Following the conferences and exhibitions, the learning
lishing the vendor-neutral submitted articles of industry continues again with ICT Today’s technology articles
subject experts and technology thought leaders to whom that are digitally archived for a couple of years, serving
we owe ICT Today’s success and distinction from other ICT as a valuable resource library to be accessed between
and data center publications. Some articles are written conferences and while BICSI standards manuals and
by the innovators of new technologies, while other articles specialty publications are updated.
are submitted by the designers, installers, consultants, and
To our essential workers, please know that you deserve
project managers who have first-hand experience with
thanks today and every day.
those new technology innovations.
Be Safe, Be Strong, Be Strategic.
Most new and emerging technologies have yet to receive
established standards, but BICSI ensures that best prac-
tices are introduced in the meantime. For example, Li-Fi,
COVER ARTICLE By Sundar Sankaran, Ph.D.
A high-speed wireless highway for tomorrow’s

fastest devices and services.
Imagine taking a nice, fast car onto a wide-open highway with plenty of
lanes and minimal traffic. For once, the ability to get to the destination,
as fast as possible, is unencumbered by the other cars on the road. That
is the experience of Wi-Fi 6E. It is like a private highway for the latest
and greatest devices and services.
6 I ICT TODAY
THE Wi-Fi EVOLUTION
Wi-Fi has come a long way since the late 1990s. The first
and second generations of Wi-Fi (802.11b and 802.11a)
were game changers when they were released; they unteth-
ered computers from the wired Ethernet connections that
were a standard of the time.
The third and fourth generations of Wi-Fi, also known
as 802.11g and 802.11n, upped the ante with new features
for the 2.4 GHz band, as well as faster speeds respectively.
With more wireless devices in the 2000s, such as smart-
phones, tablets, set-tops, and gaming consoles, these gen-
erations ushered in the mainstream adoption of Wi-Fi.
In fact, the uptick in traffic on the common 2.4 GHz
Wi-Fi band already had consumers and their devices
looking toward the faster and less congested 5 GHz
band for reprieve.
In 2014, 5 GHz Wi-Fi got a big upgrade with Wi-Fi 5
(also known as 802.11ac). This fifth generation of Wi-Fi
finally brought the 5 GHz band of Wi-Fi into the realm
of gigabit speeds (Figure 1).
In 2019, Wi-Fi 6 (or 802.11ax) went even further
by increasing general performance on both 2.4 GHz
and 5 GHz bands. It also laid the groundwork for the
next generation of Wi-Fi that would leverage the FCC’s
release of new spectrum to compose a new 6 GHz band.
EVOLUTION OF STANDARDS
2/11/54 Mb/s 54 Mb/s 600 Mb/s 6.8 Gb/s 10 Gb/s 10 Gb/s
802.11 802.11 802.11 802.11 802.11 802.11

b, a g n ac ax ax6E
1997-1999 2003 2009 2014 2019 2020s

802.11ay uses the 60 GHz spectrum to provide 20-40 Gb/s throughput. Wi-Fi 6E opens up an additional
1.2 GHz for unlicensed use in US and 455 MHz in EU.
2.4 Kb/s 9.6-100 Kb/s 2-42 Mb/s 0.1-1 Gb/s 2-20 Gb/s
1G 2G 3G 4G 5G
1980s 1990s 2000s 2010s 2020s

Al rates are peak PHY rates at maximum configuration. Actual data rates generally 50-75% of peak PHY rates. For example, TCP throughput
is 70%, 60%, and 50% of PHY rate for 11ac, 11n, and 11a/b/g, respectively.
FIGURE 1: Evolution of Wi-Fi and 5G standards. (Image: ©2020 CommScope. All rights reserved.)
8 I ICT TODAY
UNLICENSED 6 GHZ SPECTRUM than previous generations of Wi-Fi. The reason is that
In April 2020, the FCC released a 1.2 GHz plot of unli- these new lanes on the proverbial highway can only
censed spectrum—from 5.925 GHz to 7.125 GHz—to serve be used by Wi-Fi 6E devices, which will be the latest
as the basis for new wireless communications. and fastest devices available.
An analogy to illustrate this scenario involves the
This spectrum is significant for a number of reasons: highways in India. Today, on these highways, it is not
•
It is the first addition to consumer wireless unusual for drivers to encounter any number of obstacles,
frequencies in 20 years. including the usual debilitating density of other drivers,
•
It triples the amount of spectrum available foot traffic, bicycles, and narrow lanes. Much like today’s
for Wi-Fi, adding 1.2 GHz to the previous 5 GHz Wi-Fi bands, these highways are already full, and
~600 MHz of available spectrum. some, like the 2.4 GHz band, are overflowing.
•
Because it is unlicensed, its use is not restricted Additional challenges include dynamic frequency
to a single company or geography in the way that selection (DFS) and false detection issues on 5 GHz. These
5G is, for example. As a result, it is open to multiple challenges can be thought of as the highway equivalent
uses by various users throughout the country. of drivers assuming that road construction crews will
close off a lane, preemptively overburdening other lanes
Furthermore, this additional spectrum would represent to avoid it, only to realize later that there was no road
the effective doubling of lanes on the wireless highways. construction in the first place. This scenario gives a sense
of how busy, fragile and overtaxed the current Wi-Fi
THE Wi-Fi 6E STANDARD situation really is.
How does Wi-Fi 6E benefit from 6 GHz unlicensed spec- With Wi-Fi 6E, it is similar to adding brand new
trum? In short, it makes things, especially devices, much lanes, but it is much more like adding a completely sepa-
faster. The “E” in Wi-Fi 6E refers to “extended” spectrum rate parallel highway where the road surface is smooth,
and includes operation in the new 6 GHz band. This the lanes are wide, and the only things allowed are the
is not to be confused with current Wi-Fi 6 technology, fastest, most efficient, and intelligent cars that do not get
which simply denotes the sixth generation of the Wi-Fi in the way of one another. Greater spaces between cars
standard and operates over the usual 2.4 GHz and afford a greater margin of safety or reliability, and the
5 GHz bands (Figure 2). wider lanes make room for larger vehicles to carry more
As the only Wi-Fi standard to use the new 6 GHz things and deliver them more efficiently.
band, Wi-Fi 6E is in a unique position to deliver faster It is a completely different world that allows Wi-Fi
speeds, lower latency, and higher overall performance to be used in a way it has never been used before.
Wi-Fi 6E
The FCC approved 1,200 MHz of unlicensed spectrum in the 6 GHz band.
Europe and the UK are expected to release 500 MHz in 2021.
6 GHz
2.4 GHz 5 GHz 5.925-7.172 GHz [~1 GHz]
Freq.
20/40 MHz
20/40/80/160 MHz
20/40/80/160 MHz
FIGURE 2: Wi-Fi 6E approvals. (Image: ©2020 CommScope. All rights reserved.)
THE UNIQUE ADVANTAGES OF Wi-Fi 6E
Distinct advantages of Wi-Fi 6E over previous generations
of Wi-Fi include:
As the only Wi-Fi standard
Faster Real-World Speeds
Wi-Fi 6E does not deliver faster max physical layer (PHY) to use the new 6 GHz band,
rates than Wi-Fi 6, but it does achieve faster real-world Wi-Fi 6E is in a unique
transfer rates because these are dependent on factors that
it facilitates, such as wireless signal strength, associated position to deliver faster speeds,
device capabilities, and traffic from legacy devices. For
lower latency, and higher
example, while 802.11b devices will work on an 802.11ax
connection, they will not be able to fully realize the faster overall performance than
speeds associated with the new standard. Meanwhile, all
Wi-Fi 6E devices will be fast.
previous generations of Wi-Fi.
Deterministic Operation
Orthogonal frequency-division multiple access
(OFDMA) was introduced into the Wi-Fi standard via
Wi-Fi 6. It is a new way of handling traffic between technologies. This low latency enables a host
a wireless access point (WAP) and various client devices, of real-time applications over the convenience
such as cellphones, tablets, and laptops. If one thinks of a wireless connection.
of the relationship between WAPs and client devices
as a package delivery service, the old way of modulating Dedicated Channels
traffic might look something like a fleet of cars traveling Specifically, Wi-Fi 6E addresses an increasingly relevant
to and from the shipping center and various receivers, shortcoming of previous Wi-Fi generations, which is their
delivering one package at a time. inability to keep up with the demands of multiple devices
By comparison, OFDMA is a semitruck carrying vying for the same wireless bands. Wi-Fi 6E opens up
boxes destined for multiple receivers in a single trip. seven discrete 160 MHz channels (or fourteen 80 MHz
It is more efficient, reduces the overall shipping time channels) for transmission.
(or latency), and reduces traffic on the roads. The To put this in perspective, today’s 2.4 GHz and 5 GHz
wider lanes on the 6 GHz highway enable more semis bands deliver a total of 49 discrete channels, while 6 GHz
to deliver more packages. offers 109. Thus, the addition of the 6 GHz band more
The ability to improve package delivery service than triples the number of Wi-Fi channels available—as it
enables much higher levels of quality of service (QoS) goes from 49 channels to 158. These additional channels
for these connections that, in turn, can be leaned will not only help to lower congestion by keeping less
on to fulfill specific service level agreements (SLAs) demanding devices on more common channels, but
to support time-sensitive and high-bandwidth will also allow faster traffic to fly through dedicated
applications, such as gaming and telemedicine. high-speed lanes.
Additionally, the new Wi-Fi 6E standard allows the
Latency same power spectral density across all channel bandwidth,
In addition to OFDMA, minimal congestion on the thereby reducing the penalty for accessing larger channels.
6 GHz highway also allows Wi-Fi 6E to bring packet This, in conjunction with less congestion due to more
latency down to around 2 ms—much less than the spectrum and lack of legacy devices, will enable the
blink of an eye and much more in line with competing latest devices to make use of larger bandwidths.
10 I ICT TODAY
THANK YOU TO OUR SPONSORS
DIAMOND SPONSOR
PLATINUM SPONSOR
ICT FIELD OF VISION SPONSORS
bicsi.org/fall
A Wi-Fi network, due in large part to its attendant frequencies
and equipment, has a comparatively much shorter range than cellular
networks using commercial towers to transmit radio signals.
LEAD APPLICATIONS 6 GHz Backhaul

The advantages of speed, latency, and reliability make Most internet service providers install modems at the
Wi-Fi 6E purpose-built for a number of demanding edge of the home where cable or optical fiber enters the
applications: house, but this is not the best place for a router to deliver
Wi-Fi to the entire home. Oftentimes, either operators
Wire Replacement or consumers will install two or more WAPs in order
Ever since Wi-Fi took the place of a wired Ethernet to achieve consistent speeds throughout the domicile.
connection, it has struggled to deliver comparable These WAPs need to communicate with one another,
performance. If the connection is to work and work well, especially in a mesh system, in order to relay data to and
a cable has always been a simple solution. However, Wi-Fi from devices and the operator’s network. The connection
6E narrows the gap so closely that the differences between between WAPs in this system is called a backhaul, and the
wired and wireless are almost negligible. It makes Wi-Fi faster it is, the faster devices will connect through this
6E a particularly good candidate for applications in which chain to the internet.
running wires through existing structures can be both
costly and inconvenient. Dedicated Services
Some of the biggest winners from the developments
in Wi-Fi 6E will be applications, such as virtual reality
(VR), augmented reality (AR), telemedicine, and
tele-education. Each of these is incredibly bandwidth
intensive and requires extremely low latency. The Wi-Fi
6E’s ability to control congestion and boost the physical
layer’s throughput makes it particularly well-suited for
enabling the growth and mainstream adoption of these
applications and others like them.
General Improvements in Existing Experiences

A rising tide of new spectrum lifts all connected
experiences. As 6 GHz channels are carved out for
specific devices or applications, such as set-tops, tablets,
and streaming 8K video, they will arrive without the
buffering, latency, time outs, adaptive bit rate (ABR)
image quality adjustments, and many other
compromises associated with today’s Wi-Fi.
12 I ICT TODAY
Enterprises and Venues CELLULAR VERSUS Wi-Fi
With a renewed capacity to handle high-traffic scenarios There is a big debate about whether Wi-Fi 6E will
in highly dense environments, such as large corporate eventually win out over 5G. Whereas 5G offers incredible
campuses, public hotspots, multi-dwelling units (MDUs), mobility, low latencies, and deterministic performance,
concert halls and stadiums, Wi-Fi 6E can be an important Wi-Fi 6E provides more consistent and deterministic
complement to 5G networks. In fact, the divergence performance with comparatively low latency, cost,
in design objectives between Wi-Fi and 5G allows these and complexity.
otherwise competing and overlapping technologies A reasonable and measured conclusion is that they
to play off of one another’s strengths in multi-modal will both coexist and complement each other’s strengths
wireless deployments (Figure 3). and shortcomings. The more important question is how?
Most consumers are agnostic when it comes to wire-
less networks; they care only that they work when they
need them to work. Enterprises deploying these technolo-
gies, however, require a better understanding of how they
compare and potentially interoperate. As such, it is import-
ant to compare the historical tenets of cellular and
Wi-Fi networks:
Licensed Versus Unlicensed Networks

One of the primary differences between cellular networks,
like 5G, and Wi-Fi is that cellular networks are licensed.
Using licensed radio spectrum allows cellular networks
to be implicitly more controllable. For example, com-
panies that deploy cellular networks must pay for the
exclusive right to broadcast data over those frequencies.
Furthermore, in order to use those networks, clients must
use SIM cards that confirm that the user has permission
to gain access to a given network.
On the other hand, Wi-Fi networks run on unlicensed
spectrum. On these networks, anyone can develop and
deploy devices that run on supported frequencies. Tradi-
tionally, these have included the two bands discussed
earlier: 2.4 GHz and 5 GHz. This also allows for a certain
universality in access. For example, a Wi-Fi device can
be used on a Wi-Fi network anywhere in the world,
whereas cellular networks require validation (through
an associated SIM) to attach to each network.
These very different uses of spectrum create
commensurately unique approaches to the security and
development of associated devices, as well as the inherent
capabilities of those networks. For example, a Wi-Fi net-
FIGURE 3: Himachal Pradesh Cricket Association Stadium
work, due in large part to its attendant frequencies and
in Dharamsala, India. Bottom photo shows the Dharamsala
Cricket Stadium’s dense wireless access point deployment. equipment, has a comparatively much shorter range than
The Dalai Lama actually lives nearby. cellular networks using commercial towers to transmit
radio signals. These fundamental differences in design the ability of each technology to cater to evolving
have historically created a divergence in the applications global demands.
of these technologies; Wi-Fi is normally used indoors,
whereas cellular is normally used outdoors. Wi-Fi 6E AND 5G
However, there are many other considerations In addition to the traditional tradeoffs of cellular versus
involved in comparing these two general approaches Wi-Fi, the evolution to 5G and Wi-Fi 6E has introduced
to wireless connectivity: new features and capabilities that further blur the lines
between the two technologies. In fact, as cellular and
Cost: Because the licensing of cellular technolo-
• Wi-Fi have evolved, they have also begun to converge
gies makes cellular infrastructure more expensive and share innovation. Some examples include OFDMA,
for clients, Wi-Fi has a distinct advantage in terms multi-user MIMO (MU-MIMO), and beamforming, which
of deployment costs. Also, Wi-Fi typically avoids are used across modern forms of both networks
the subscriber costs associated with paying to manage growing data and device traffic and
a network operator to use its infrastructure to improve transmission rates.
and licensed spectrum. The FCC’s release of 6 GHz unlicensed spectrum,
in addition to boosting the capabilities of Wi-Fi, also
Ubiquity: Wi-Fi is everywhere. With its backwards
• provides a new outlet for 5G to offload traffic. Even
compatibility, its evolving infrastructure ensures though cellular networks traditionally run on licensed
that investments in legacy devices are protected. spectrum, they can also use unlicensed spectrum
This contrasts with cellular chipsets, which usually to supplement bandwidth and coverage.
support only the latest generations of devices. The When Wi-Fi 6E and 5G are compared, there are some
result is that there are billions more Wi-Fi devices very obvious similarities and advantages to each approach:
in operation than cellular devices.
Speed: Wi-Fi 6E supports a peak data rate of 9.6
•
Ease
• of Deployment: The two technologies’ Gb/s, which is comparable to 5G, but Wi-Fi 6E also
divergent allegiances to licensed and unlicensed achieves a spectral efficiency of 62.5 bps/Hz, more
spectrum mean very different levels of speed and than double that of 5G’s specified 30 bps/Hz.
complexity in deployment. Wi-Fi can be set up
in minutes by most consumers, whereas cellular Delivery: Both standards are capable of supporting
•
still requires skilled administration. the latest and most demanding AR, VR, and
IoT applications (e.g., telemedicine) through
Coverage
• and Mobility: Because of the aforemen- high-speed data rates and extremely low latency.
tioned differences in design objectives between Both also support new protocols to extend the
cellular and Wi-Fi, cellular has evolved to deliver battery life of associated client devices.
wider outdoor coverage that can handle highly
mobile devices. Wi-Fi thrives indoors but is most Density: This is one area where Wi-Fi unequivo-
•
effective for short-range connectivity. cally comes out on top. Wi-Fi 6E works very well
for extremely dense environments, such as stadi-
Cadence
• of Innovation: Historically, Wi-Fi has ums, campuses, and large venues. A single Wi-Fi
evolved faster than cellular via a technology 6E WAP, for example, can serve up to 1,024 clients
refresh cycle every five years versus cellular’s concurrently. Its trigger frame feature, which
10-year cadence. As bandwidth use continues is related to OFDMA, enables scheduled access
to grow and as demanding new applications and results in improved transmission, bringing
take hold, these innovation cycles will dictate its reliability in line with 5G’s.
14 I ICT TODAY
the use of PoE to reduce faultfinding time, especially
In addition to the traditional in far-flung locations. With a non-PoE system, a power
failure event requires the on-premise attention of an elec-
tradeoffs of cellular versus trician to investigate the root cause of the outage. With
Wi-Fi, the evolution to 5G and a PoE-fed Wi-Fi 6E WAP, power and data are combined
and centralized at the network switch in an equipment
Wi-Fi 6E has introduced new room with dedicated power circuits, thus simplifying and
features and capabilities that automating the faultfinding process. This reduces the time
it takes to track down and repair outages, significantly
further blur the lines between improving the mean time to recovery (MTTR).
the two technologies. THE Wi-Fi 6E OPPORTUNITY

The combined advantage of Wi-Fi 6E’s new performance
characteristics and the technology’s favorable deploy-
Security: Today, Wi-Fi roaming and cellular roam-
• ment and management costs make it a very competitive
ing are relatively seamless, and the latest Wi-Fi choice for indoor and enterprise applications. Its ability
security updates have made it as secure as cellular. to leverage 6 GHz spectrum will create further improve-
These protocols allow for individualized encryp- ments to latency and reliability for dedicated services and
tion that potentially protects users from hackers. applications, as those are further carved out in the
coming years.
Although consumer devices, such as cellphones and The bottom line is that the future of Wi-Fi is prom-
tablets, have evolved to include both cellular and Wi-Fi ising—both as the evolution of an incredibly successful
radios, enterprises considering how and when to imple- wireless networking standard as well as a complement
ment one or both technologies should rely on a detailed to cellular technologies like 5G. Wi-Fi 6E is made for the
assessment and plan. They should also work hand-in- 10 Gb era. It is an opportunity for operators and enter-
hand with vendors who understand both technologies prises to usher in a new age of exciting new wireless
and how they compare and complement one another. experiences within the home and beyond.
PoE, CAT 6A, AND Wi-Fi 6E AUTHOR BIOGRAPHY: Sundar Sankaran, Ph.D., is a VP
Power over ethernet (PoE) is the preferred technology for of Engineering for CommScope’s RUCKUS portfolio where
delivering power to newer edge devices and WAPs. The he leads the access point hardware and software engineering
latest 802.3bt PoE standard (also known as 4-Pair PoE teams. He is a wireless industry veteran with extensive experience
or simply 4PPoE) stipulates support for a full 90 watts designing wireless systems. Sundar joined Ruckus Networks from
at the power source equipment (PSE), which is deliver- Atheros/Qualcomm where he was a senior director of technology
able via Cat 6A cabling. Although older WAPs tend to and served as the overall engineering lead. His responsibility
draw a minimal amount of power, Wi-Fi 6E WAPs will was to deliver silicon along with reference hardware and software
likely require more power to drive all their radios and on multiple Wi-Fi chip programs. Prior to Atheros/Qualcomm,
provide power for devices connected via their USB ports. he had stints at Intel, ArrayComm, and Infosys. He is a co-inventor
As such, IT departments upgrading older WAPs to Wi-Fi on 18 US patents, as well as several pending patents. Sundar
6E WAPs will also deploy Cat 6A cabling, which supports earned a Bachelor’s degree in electronics and communication
transfer rates of up to 10 Gb/s to prevent network bottle- engineering from Anna University and Master’s and Ph.D.
necks and fully support new PoE demands. degrees in electrical engineering from Virginia Tech. He can
Mission-critical use cases that involve the use of Wi-Fi be reached at sundar.sankaran@Commscope.com.
6E WAPs and connected edge devices will also demand
Li-Fi: The Future
of LED Lighting
Technology
By Miro Sino, RCDD, P.Eng, LC, LEED AP
Most ICT professionals are probably aware of how LED lighting technology has taken the industry by storm
since 2013, especially with power over Ethernet (PoE) applications. LED lighting is outpacing what fluorescent
lighting was to energy efficiency in the 1980s. A 100 watt incandescent lamp was replaced with a 32 watt
compact fluorescent lamp; those two are now replaced with an 8 watt LED lamp.
The LED lighting industry continues to evolve so rapidly that lighting standards on light level output, LED
drivers, and LED lamp performances are continually updated. Undoubtedly, LED lighting has become the promise
of the future, allowing designers and engineers to comply with leadership in energy and environmental design
(LEED) guidelines for energy efficiency and to meet strict ASHRAE guidelines for reducing energy usage in new
and existing buildings.
16 I ICT TODAY
The most recent and ingeniously innovative use of LED
lighting is in its application for data transmission. This
new technology that is showing great promise for the
future of wireless data transmission is called light
fidelity (Li-Fi). It is committed to making an internet
connection as simple as turning on a light switch!
UNDERSTANDING THE TERMINOLOGY

Light fidelity is a visible light communication (VLC)
system and a wireless communication technology utiliz-
ing light in high frequency to transmit data to and from
devices through light output. The VLC term dates to the
1880s. It includes any use of the visible light portion
of the electromagnetic spectrum to transmit informa-
tion. The term VLC was further defined by the IEEE
802.15.7r1 standardization committee in January 2016.
The Li-Fi concept was invented by Professor Harald
FIGURE 1: With each light or bulb serving as an internet
Haas from the University of Edinburgh, Scotland in 2011 connection and a means to gather data, future innovations
when he demonstrated for the first time that by flickering and applications may be limited only by the imagination.
the light from a single LED, he could transmit far more
data than a cellular tower.1 The term Li-Fi was launched (the signal) as data (Figure 1). This detector is incorpo-
at his 2011 TED Global Talk where he introduced the idea rated in a communications device.
of wireless data from every light bulb; his talk has been The high frequency of light/data transmission
watched online more than 2.4 million times. is undetectable to the human eye. The LED lights used
He remains Professor of Mobile Communications to transmit Li-Fi signals are modulated at such a fast rate
at the University of Edinburgh and the co-founder that the eye cannot perceive the modulation or flicker.
of pureLiFi along with Dr Mostafa Afgani2 where This is similar to the way eyes do not interpret the break
research and modifications toward standardization between film frames in a motion picture. Just as people
are ongoing. see a smooth motion on the cinema screen, they will see
an uninterrupted source of steady light streaming from
THE CONCEPT OF Li-Fi a Li-Fi-enabled luminaire. As a comparison, the lowest
AND HOW IT WORKS frequency at which the lights are modulated is 1 MHz,
“Light fidelity is a high-speed, bidirectional networked which is 10,000 times faster than the refresh rate of com-
technology achieving mobile communication of data puter screens. Since the intensity of modulation cannot
using light. Li-Fi is comprised of multiple light bulbs that be seen, the data communication is the same as other
form a wireless network.” 3
radio systems. This allows users to be connected to wire-
When an electrical current is applied to an LED light less internet where there is Li-Fi-enabled light.
bulb, a stream of light or photons is emitted through the Because LED lamps are electronic devices, their
bulb. Each tiny LED lamp is a semiconductor device, currents and, therefore, their lights can be modulated
which means that the brightness of the light flowing at extremely high speeds. The modulated light is detected
through it can be changed at an extremely high speed. by a photosensor in the electronic device that receives
This allows a signal to be sent by modulating the light it, which then converts the pulsing light into an electri-
at different rates. The signal can then be received by cal current through a transceiver. This electrical current
a detector that interprets the changes in light intensity forms the data signal.
The high frequency of light/data transmission is undetectable
to the human eye. The LED lights used to transmit Li-Fi signals
are modulated at such a fast rate that the eye cannot
perceive the modulation or flicker.
Li-Fi's OPERATIONAL LIGHT SPECTRUM caused by natural disruptions in daylight or sunlight.

For Li-Fi technology, communication data information Since sunlight is constant light and Li-Fi technology
is provided through the following spectrum ranges: modulates the light at very high rates, the sunlight can
•
Visible light wavelengths from 380 nm to 740 nm be filtered out at the receiver. Outdoor receivers have
(430 to 770 THz light frequency) been tested under 77,000 lux of sunlight (or 7,700
•
Infrared light wavelengths from 700 nm to 1 mm foot-candles [fc]).4 As a reference, a brightly lit office
(430 THz to 300 GHz light frequency) is normally at 500 lux (50 fc).
•
Ultraviolet light wavelengths from 10 nm to If all power to a light is turned off, then there is no
400 nm (8 × 10 Hz to 3 × 10 Hz light frequency)
14 16
Li-Fi. However, Li-Fi technology can be enabled to low
enough light levels in a room that make it appear dark
Because UV light has damaging effects on human and still transmit data. There is consistent performance
health, only visible and infrared light spectrums are used between 10 and 90 percent illumination.5
in Li-Fi technology. This opens up 10,000 times greater Communications with Li-Fi’s technology has been
bandwidth than existing Wi-Fi radio frequencies (RFs) achieved at light levels down to 60 lux (6 fc). As a com-
on the visible spectrum only. Therefore, visible light parison, the IESNA standard light level for reading
spectrum is predominantly used. Options are also being is between 200 and 500 lux (20-50 fc). Lighting along
explored using the invisible infrared light spectrum, since a perimeter pathway around a building at nighttime
it is already being used to send information (uplink) back is around 5 to 6 fc.
to the LED light bulb.
Because light bounces off surfaces and light reflected Li-Fi SPEED
from surfaces also carries data, Li-Fi is not strictly a line- As a dense and fast wireless communication technology,
of-sight technology. Rather, Li-Fi is a cellular communi- Li-Fi offers greater speed and density than RF solutions.
cation system, and the data rate is not dependent on the The following notable achievements exemplify
line of sight but on the signal quality at the device. Signal its evolution:
quality can be defined by the ratio of the desired data •
August 2013: Data rates of over 1.6 Gb/s were
versus any interfering data and noise. demonstrated over a single-color LED with
It is also important to understand that Li-Fi an upstream communication rate of about
is a bidirectional wireless communications technology 600 Mb/s.6
that allows high speed transmission in both uplink and •
April 2014: A company from Russia developed
downlink simultaneously. a Li-Fi wireless local network module transferring
data at 1.25 Gb/s, while striving to achieve
LIGHT LEVELS FOR Li-Fi up to 5 Gb/s.7
Light fidelity can operate in daylight and even in direct •
In 2014, a new record was established by
sunlight conditions, as the modulated light can still be a company in Mexico at transfer data speeds
detected. It relies on detecting the fast changes in light of up to 10 Gb/s across the light spectrum
intensity and not on the absolute or slowly varying levels emitted LED lamps.8
18 I ICT TODAY
•
January 2020: Li-Fi was presented at the Li-Fi VERSUS WI-FI AND CELLULAR
Consumer Electronics Show (CES) in Las Vegas. “Li-Fi seeks to resolve the global struggle for diminishing
It was announced that ultra-high speed Li-Fi wireless capacity resulting in the spectrum crunch. By
communication technologies for consumer offering secure, reliable, high-speed communication
applications are available. The LaserLight networks, the Li-Fi technology seamlessly integrates
Li-Fi communication technology, for example, data and lighting utility infrastructures to significantly
achieves up to 20 Gb/s high-speed data rates. reduce energy consumption.”10 Li-Fi claims and
technology distinctions include:11
Upload transmission speeds are at about 60 percent •
Li-Fi is a wireless technology that holds the key
the rate of download speeds. to solving challenges faced by 5G. Li-Fi can trans-
mit at multiple gigabits, otherwise known as fiber
REAL LiFi optic speeds. It is more reliable, virtually interfer-
The term “Real LiFi” was introduced to describe a lighting ence free, and uniquely more secure than radio
system that is fully networked, meaning all the lights technology, such as Wi-Fi or cellular.
on a Li-Fi network can speak to each other, thus provid-
ing a continuous network and handover data as users •
Radio frequency is vulnerable to interference from
move between lights and around the space (Figure 2). a wide range of devices, such as cordless phones,
Each light has its own unique IP address, which offers microwaves and neighboring Wi-Fi networks. Li-Fi
location tracking and advanced geofencing that can be signals can be defined by the area of illumination,
deployed within the network. which means interference with radio-frequency
Real LiFi is bidirectional, full duplex, and high- devices is much simpler to avoid and even stop
speed. Data is received and transmitted at the same altogether. This also means Li-Fi can be used
time, resulting in a fast user experience, especially when in RF hostile zones, such as hospitals, airplanes,
streaming video and using virtual and augmented reality. and power, petrochemical and nuclear plants.
FIGURE 2: “Multiple light bulbs form a wireless network covering arbitrarily large areas wherever there is light. With Real LiFi,
you can move between lights, download videos and make Skype calls.”9
•
Radio frequency communication requires radio At a minimum, ICT designers and installers should
circuits, antennas and complex receivers, whereas keep in mind the differences between Li-Fi and Wi-Fi
Li-Fi is much simpler and uses direct modulation as illustrated in Table 1.
methods similar to those used in low-cost infrared
communications devices, such as remote control Li-Fi SECURITY CONSIDERATIONS
units. LED light bulbs have high intensities and The active area of Li-Fi is confined to the area where light
therefore can achieve very large data rates. is distributed (where light touches). The signal remains
hidden to anyone outside this area. The light is controlla-
•
Li-Fi can deliver multiple Gb/s speeds in mobile ble by shutting the door or pulling the blinds to protect
devices. This next generation technology will drive the network. This makes it more secure and easier to con-
wireless beyond any current capability, opening trol access to a network. As long as transparent materials,
up unprecedented bandwidth. such as windows, are covered, access to a Li-Fi channel
is limited to devices inside the room.14
•
Light can be contained and secured in a physical Since light cannot penetrate through walls, accessing
space. Li-Fi enables additional control as it offers data from a light that is in the other room is not possible.
precise localization for asset tracking and This may be an inconvenience, but it also presents a great
user authentication. security feature. According to pureLiFi, existing security
protocols for encryption and authentication can be
Interestingly,“Li-Fi would not work if the device leveraged in Li-Fi systems to provide even better secured
is in your pocket or bag, unless there is a transparent wireless systems.
cover. Li-Fi is a complimentary technology that works
alongside other wireless technologies such as Wi-Fi. APPLICABLE STANDARDS
If the light signal to a Li-Fi enabled device is below the TO Li-Fi INSTALLATIONS
receiver’s threshold, then it will not receive data. In that Standards applicable to the installation of a Li-Fi system
instance, radio systems or cellular networks, if available, would be those relevant to the installation of telecom-
will continue to deliver data. However, the moment munications and structured cabling systems, as well
the device begins to receive light from a Li-Fi light bulb as all electrical standards applicable to the installation
[or luminaire], then the device will resume high speed of lighting and energy efficiency.
communications using light as an additional Until the Li-Fi technology is firmly established,
communications medium.” 12
the following standards, including but not limited
Furthermore, “Li-Fi currently offers latency by to, shall be complied with for all Li-Fi installations.
a factor of three times lower than Wi-Fi and can radi- Close coordination with the authority having
cally enable innovation, automation, and applications, jurisdiction (AHJ) inspectors and plan reviewers
such as virtual reality and augmented reality.” 13
will be necessary.
Li-Fi Wi-Fi
Operation Uses LED Uses modem
Data Transfer Speed >1 Gb/s About 150 Mb/s
Security High Low
Cost Lower, visible light spectrum is free Higher, radio spectrum requires license
Power Consumption Potentially Less More
TABLE 1: Essential differences between Li-Fi and Wi-Fi.
20 I ICT TODAY
1. BICSI, Telecommunications Distribution Methods 8.
IEEE 802.11bb - Light Communication Standard
Manual (TDMM) 14th Edition 9.
National Electrical Code (NEC)/Canadian Electrical
2.
ANSI/BICSI 007-2020, Information Communication Code (CEC) and Electrical Safety Branch
Technology Design and Implementation Practices Regulations and Bulletins, latest editions.
for Intelligent Buildings and Premises 10.
State/Provincial Building Codes, latest editions.
3.
Commercial Building Communication Cabling 11.
AHJ Bylaws, latest editions.
Standard ANSI/TIA-568.1-E for Cat 5e and higher. 12.
IESNA 10th edition guidelines.
4.
Commercial Building Standards for Telecommuni- 13.
ASHRAE 2016 edition guidelines.
cations Pathways and Spaces (ANSI/TIA-569-E). 14.
LED drivers shall be UL/CSA approved and meet
5.
Grounding and Bonding Requirements for requirements of the NEMA and Federal Communi-
Telecommunications in Commercial Buildings cations Commission Rules and Regulations, Part 18.
(ANSI/J-STD-607 and CSA T527-94). BICSI’s TDMM, 15.
LED drivers shall be manufactured in accordance
14th Edition, Chapter 8, Bonding and Grounding. with ANSI C82.11.
6.
UTP End-To-End Performance Testing for Balanced
Twisted-pair Cabling (ANSI/TIA-1152-A). For international standards, it is advised to consult
7.
Administration Standard for Telecommunications ISO/IEC standards also; applicable standards for telecom-
Infrastructure of Commercial Buildings munications cabling, pathways, and spaces should
(ANSI/TIA-606-C). be followed.
Factory Preloaded
Fiber Optic Solutionists
Fiber Instrument Sales, Inc. Fiber Interconnects
48 Hour Turnaround On Select Preloaded Interconnects
Enclosures Ship Fully Loaded With Adapter Plates,
Splice Trays, and/or Pigtails
∙ Slide-out Rack Mounts ∙ Deluxe Wall Mounts
∙ SC/LC Configurations ∙ Singlemode/Multimode Options
Advantages
∙ Ships in 48hrs
∙ Plug-n-Play Application
∙ Reduced Field Installation Time
∙ No Minimum Order Quantities
* Lead times are subject to change 1, 2, & 4 RU Slide-Out Rack Mounts Economy and Deluxe Wall Mounts
Sales Representatives Are Waiting For Your Call!

www.fiberinstrumentsales.com
1-800-5000-FIS(347) FOR OUR FULL LINE OF PRODUCT OFFERINGS
INSTALLATION OF Li-Fi SYSTEMS
In order to enable Li-Fi service in any building or exterior
application, the utility front end internet service equip-
The highest quality of lighting
ment is still required. Once service is provided at a build- controls can be achieved
ing and the high-speed connection is established, the
Li-Fi network can be utilized to distribute internet service
by creating a network
by using PoE or power line communications (PLC). of addressable lighting.
All Li-Fi compatible drivers that power LED lighting
need to be connected via minimum Cat 6A cabling or
optical fiber to the head-end telecommunications equip-
ment (e.g., patch panel, router, modem) depending on Some projects completed in Europe indicate numbers
the number of lights that will provide the data signals as high as 25 percent reduction in overall cabling installa-
to the drivers, which will then transmit the information tion time and a “50 percent installation cost savings than
to users in high-frequency. the equivalent traditional installation.”15
Category 5e and higher cables may also be used for The highest quality of lighting controls can be
the cabling structure between the utility provided signal achieved by creating a network of addressable lighting.
equipment and the LED lights. However, the recommen- In order to accomplish this, minimum Ethernet Cat 5e
dation in TIA TSB-162 and the minimum requirement cable or higher has to be run to each luminaire, although
in ANSI/BICSI 008-2018, Wireless Local Area Network Cat 6A or higher cabling is recommended for achieving
(WLAN) Systems Design and Implementation Best Practices the higher uplink speeds of at least 10 Gb/s.
is one Category 6A cable. Optical fiber cabling (OM3 The benefits of addressable lighting include the
or higher) or two balanced twisted-pair cables (Category ability of creating lighting zones, which in turn allows
6A or higher) are recommended to support speeds higher for creation of lighting scenes, energy savings, and ease
than 10 Gb/s. of troubleshooting. If lighting in general is also used
to provide Li-Fi service, all cabling to devices become
WIRING INSTALLATIONS obsolete, unless the device is located outside of the
FOR Li-Fi NETWORKS Li-Fi light range where communication outlets may
There are three solutions for providing the cabling be provided. For the most part, devices tend to be
infrastructure for Li-Fi wireless communications: in illuminated areas where people are able to see
•
Currently, PoE is the optimal solution for driving while they operate them.
the network cabling structure and, therefore, the Based on this, the projects that would likely reap
preferred option for new installations. the reduction in cabling installation time and overall
costs similar to the European projects would be those
•
Power line communications is the preferred option that originally intended to install addressable lighting.
for retrofit installations. PLC can be used for Adding addressable lighting to a project can be very
sharing the available capacity between lamps expensive. It is mainly used in large projects where
attached to any given circuit ring. lighting addressability brings real value to daily opera-
tions, such as at convention centers, theatres, hospitals,
•
The “Cloud RAN” architecture can also help airports, and large office buildings where new and different
alleviate the pressure on the cabling infrastructure. tenants need a different square footage almost every time
they move in. With non-addressable lighting, wiring
However, the long-term cabling infrastructure modifications are required, which could make the tenant
is resolved by using PoE in new offices, which is proving improvement costs expensive. Addressable lighting would
to be more efficient for smart buildings and lighting. reduce this process to programming only.
22 I ICT TODAY
However, what happens to a building that did CURRENT CHALLENGES
not intend to have addressable lighting due to budgetary TO Li-Fi TECHNOLOGY
reasons? For starters, the Wi-Fi service is always there. As with most emerging technologies, there are challenges
Secondarily, depending on the number of devices that to overcome, and Li-Fi is no exception:
are wired into the network, an engineer or designer would 1.
Because the technology is in its early stages,
have to weigh the costs of each option to determine the Li-Fi equipment is not yet widely available. There
feasibility of adding addressable lighting with Li-Fi service are a few sources where some equipment, such
and its added benefits. There are certainly other consider- as transceivers, lighting, LiFi compatible drivers,
ations in that decision-making process, such as future routers and kits, can currently be purchased
needs of the space, health implications, interference with (e.g., Oledcomm, pureLifi, Signify, VLNComm).
other systems and equipment, the client’s knowledge
of Li-Fi, and potential future savings. A life cycle cost 2. A Li-Fi transceiver is needed as an add-on for
analysis may be required because lighting energy savings mobile phones, tablets, computers and other
may also offset the initial capital cost with a rate electronic devices. Large manufacturers have
depending on the size of the building. yet to incorporate a Li-Fi transceiver into their
Figure 3 is a diagram of a typical Li-Fi installation. mobile devices.
This diagram is provided for guidance only, since actual
installations would have to be evaluated on a project 3. Light emitting diode luminaires that work with
basis by the engineer of record. Li-Fi technology are currently limited. Lighting
FIGURE 3: An example of a Li-Fi installation.
manufacturers have yet to catch up with manufac-
turing Li-Fi compatible drivers and to make 0-10V
dimming drivers, for example, a standard option A Li-Fi transceiver is needed
available to lighting designers and electrical engi-
neers. Regular LED drivers generally used for pro-
as an add-on for mobile phones,
viding lighting do not have the capability to create tablets, computers and other
electronic devices.
a Li-Fi network. Li-Fi LED drivers are specially
designed to emit light at high frequencies required
for Li-Fi transmission. Similar to the computerized
devices needing a Li-Fi optic transceiver in order
for the Li-Fi technology to succeed at a mass level,
it is imperative that the lighting manufacturers global light communications standard is IEEE 802.11bb.
globally adopt this technology and offer the Li-Fi Its task group, with input from across the Wi-Fi ecosystem,
driver as an option. is aiming to deliver an official ratified 802.11bb standard
by mid 2021.
4. Data or ICT installers need to become familiar Although Li-Fi is most akin to 802.11ad, which
to a certain extent with lighting terminology and is higher transmission rates at smaller distances (typically
installations. Electricians, traditionally installing room size), rather than the 802.11ac and 802.11ax for
line voltage systems such as lighting, will need LAN over distance, the standard is yet to be established.
to gain an understanding of telecommunications In addition, Li-Fi technology is being used more and
systems and connectivity, such as cabling into more in recent seminars held by IESNA, although the
a luminaire with Li-Fi compatible drivers. Large terminology is yet to be incorporated into the IESNA
contracting companies that have installers trained official lighting handbook.
in both types of installations are in the best Until there are more standards that may ignite
position to make a smoother transition toward further demand for Li-Fi, electrical engineers, telecommu-
future Li-Fi installations. nication/ICT designers, installers, and lighting designers
need to familiarize themselves with the Li-Fi technology
5. Municipalities are not yet familiar with this tech- and the equipment required for a successful installation.
nology and its requirements. Some municipalities
have requirements that forbid any telecommuni- FUTURE POTENTIAL
cation circuits to be interconnected in any way “It is predicted that future home and building automation
to lighting circuits. For installers wishing to pursue will be highly dependent on the implementation of Li-Fi
Li-Fi installations in their municipalities, it may be technology for being secure and fast.”16 Because light
important to present a comprehensive description cannot penetrate through walls, for example, the signal
regarding the use of Li-Fi that highlights the cannot be hacked from a remote location. According
safety of the installation. to Research and Markets, the global Li-Fi technology
market is expected to reach $35.82 billion by 2028,
WHAT IS NEXT? growing at a compound annual growth rate (CAGR)
International standards body IEEE is already working of 71.20 percent during 2018-2028.
with Li-Fi technology by establishing definitions to pro-
vide its dedicated place in the world of ICT. Like Wi-Fi, Potential applications and use cases include:
Li-Fi is wireless and uses similar 802.11 protocols. The •
On the road vehicle communication
objective is to extend 802.11 to include light as a medium •
Remotely operated underwater vehicles
for wireless communications transmission. The proposed (ROVs) up to 200 meters depth
24 I ICT TODAY
•
Roadway lighting REFERENCES:
•
Retail for direct shopping 1. “Li-Fi Has Just Been Tested in The Real World, And It's 100
•
Hydro companies (meter reading) Times Faster Than Wi-Fi,” Science Alert,24 November 2015.
•
Green design rating systems (e.g., LEED, Passive 2. Harald Haas, LinkedIn
House, Living Building Challenge) could include 3. Pure WiFi website, https://purelifi.com/"
credits for different measures taken for creating https://purelifi.com/
greener and healthier building designs that are 4. Scace, Sarah. “Three Li-Fi Misconceptions,” pureLiFi,
also more energy efficient 14 May 2017.
•
Hospital and industrial industries 5. Ibid.
•
Live streaming, as well as virtual and augmented 6. “Pure VLC Demonstrates Li-FI Streaming Along
reality applications with Research Supporting World’s Fastest Li-Fi Speeds
•
Military applications and operations up to 6 Gbps,” Press release. Edinburgh, Business Wire,
6 August 2013, Li-Fi Wikipedia.
Right now, lighting acts as a burden on the data 7. Pototsky, Dan. “Russian Firm’s Li-Fi Internet Solution
system requiring cabling to each light only for the benefit Winning Foreign Clients,” Science & Tech, 30 June 2014.
of lighting controls with some energy savings. However, 8. Vega, Anna. “Li-Fi Record Data Transmission of 10Gbps Set
with Li-Fi technology, LED lights would become more Using LED Lights,” E&T Magazine, 14 July 2014.
of an active member by also providing data through 9. Scace, Sarah. “Three Li-Fi Misconceptions,” pureLiFi,
lighting while promising significant energy savings. 14 May 2017.
The possibilities of how that data can be used in the 10. pureLiFi website.
use cases described are exciting and endless. 11. Ibid.
12. Scace, Sarah. “Will Li-Fi Work in My Pocket,” pureLiFi,
AUTHOR BIOGRAPHY: Miro Sino, RCDD, P.Eng, LC, 12 June 2017.
LEED AP is the president of Sino Engineering Ltd. established 13. “The Lowdown on Li-Fi,” Enlightenment Magazine,
in Vancouver, BC in 2005. After almost two decades of working 3 June 2019.
in the construction industry in Vancouver, BC, Miro brings her 14. “Li-Fi Technology,” Techminions, 26, June 2019.
combined experience of working with lighting and telecommuni- 15. Serafimovski, Nikola. “The Future of Wireless is LiFi,”
cation systems in various types of buildings, including institutional, ISE Magazine, May 2016.
commercial, retail, residential and industrial settings. Her experi- 16. “What exactly is LiFi?” LiFi.co,
ence with these two systems, combined with her knowledge of https://lifi.co/what-is-lifi/.2020.
lighting controls and the various communication system struc-
tures, has allowed her to understand the interconnection of these
systems when presented in the form of Li-Fi technology. Miro
obtained her RCDD and has been a member with BICSI since
2005. She obtained her Lighting Certified (LC) credential and has
been a member with IESNA since 2007. Furthermore, she is
a professional engineer in British Columbia and attained the LEED
BD+C credential in 2004. Miro graduated in 1998 at the Poly-
technic University of Tirana, Albania where she also taught after
graduation as an associate professor. She immigrated to Canada
in 2001. Miro can be reached at miro@sinoengineering.ca.
LEVERAGING INFRASTRUCTURE
TO OPTIMIZE THE EDGE
Moving Compute Functions to the Edge Makes Infrastructure More Critical
By Brian L. Kelly and The number of connected devices is growing. That rate of growth varies widely
Mark Dehmlow depending on the organization doing the projecting. The Telecommunications
Industry Association (TIA) touts on its website a number of 29 billion globally
by 2022. In a recently published paper from the TIA’s Edge Data Center Working
Group, TIA Position Paper: Edge Data Centers, that group suggests that machine-
generated network traffic could dwarf current internet growth with a forecast
of one million new internet of things (IoT) devices sold every hour by 2021.
26 I ICT TODAY
“The next wave of technology innovation is already as manufacturing, healthcare, and finance are among
here with new applications transforming the way we those that are adopting edge networks today.
live, work, and travel,” TIA says in its paper. “The huge Regardless of the application or the reason for
adoption of these new services drives exponential growth the transition to the edge, there are ways to leverage
in the total demand for data. We must provide more data infrastructure that impact the deployment, monitoring,
capacity and higher computing speeds if we hope to keep management, and operation of an edge data center
up. The sheer scale and scope of the gap we face demands (EDC). These infrastructure choices can help those
that we rethink the way we have traditionally organized responsible for the data center to address common
the design and deployment of networks and data centers. challenges related to security, environmental condi-
As many hands make light work, deploying many smaller tions, and reliability, thereby ensuring uptime for
distributed data centers seems the most viable solution.” the network and those systems that are dependent
These smaller distributed data centers, which com- on the network.
plement the cloud, are what the data center industry
today recognizes as the edge. EDGE COMPUTE EXPLAINED
Edge compute pushes computing functions closer At its core, edge compute is a shift from a centralized
to the end use, proximate to where data is generated, compute model to a hybrid distributed compute model.
analyzed, and stored. This evolution of the network real- What does that mean? For the past 20 years or so, the
izes numerous benefits, chief among them are reduced compute model has been almost strictly centralized.
latency between the end use and the compute resource On-premise data centers, cloud compute, and hybrid
(cloud or servers), improved application performance, cloud models are all centralized compute models, mean-
and optimized transmission costs. ing all the data is processed and stored in a central loca-
Applications for the edge are far-reaching. The tion (e.g., data center, cloud data center). Edge compute
fastest growth is happening in applications that are most supplements that model with additional compute
sensitive to network latency, require enhanced security resources at the edge or in closer proximity to the
or are bound by regulatory requirements. Industries such end use (Figure 1).
Centralized Compute Architecture Distributed Compute Architecture

FIGURE 1: Edge compute is a shift from a centralized compute model to a distributed compute model.
CDN: THE ORIGINAL EDGE content will suffer from high latency, potentially 220 ms.
COMPUTE APPLICATION This will appear as buffering, lagging or pixilation.
A content delivery network (CDN) is used every time If a CDN is present, with a PoP in or near Chicago, the
someone watches an online video, downloads some- content downloads quickly and consistently with the
thing from the internet, and accesses content via social latency potentially at 10 ms. The end user experience
media. With a CDN, content is uploaded to an origin is much better, and the content provider benefits from
server, which then distributes that content to all the reduced costs due to reduced bandwidth. It is a win-win
points of presence (PoP) in the CDN. Once distributed, solution for all parties.
end users pull the content from the closest CDN PoP,
rather than the origin server (Figure 2). OPTIMIZING EDGE DEPLOYMENTS
Consider this example: An end user near Chicago WITH INFRASTRUCTURE
wants to download content housed on a server Locating compute functions at the edge comes with
in London with no CDN. The downloading of that a unique set of challenges for IT and ICT managers:
• Remote and geographically
distributed sites
•
Lack of dedicated on-site IT
and ICT personnel
•
Non-traditional IT spaces with limited
security and environmental controls
Edge deployments are no different than

other data centers where operators expect
consistent uptime. Additionally, if IT or ICT
personnel are required to travel to the site
to address network issues, the price tag associ-
ated with that travel and the extended down-
time to allow for travel can quickly get out
With CDN
of hand. This makes it imperative to have
systems in place that help to ensure reliable,
long-term, and hands-off operation.
The infrastructure selected for an edge
deployment can help address these challenges.
Specifically, infrastructure serves as building
blocks that address consistency, reliability,
security, and remote management.
Consistency: Edge is not a single

•
instance but a multitude of instances
across a broad geographic landscape.
Consistent preconfigured designs
enable efficient deployments across
a multi-site distributed architecture,
FIGURE 2: A content delivery network (CDN) puts content closer as well as ongoing maintenance
to the user and is one of the first uses of edge computing. of that network.
28 I ICT TODAY
Preconfigured enclosure or cabinet offerings can
Specifically, infrastructure range from a simple package that includes a few stan-
dard components to one that is specified by the end user
serves as building blocks and built according to that spec, so it is ready to drop
that address consistency, into a remote site, plug in, and go. When setting up mul-
tiple sites, this step can save considerable time and money
reliability, security, and in labor costs. Although the organization will pay more
remote management. up front for the convenience of having the enclosure pre-
configured, time and labor usually are reduced during
installation. If a technician or installer is required to travel
to the site to oversee or complete the installation, travel
Reliability: High quality, standards-based solutions
• and perhaps other expenses need to be considered
are a critical first step to ensuring site reliability as well.
where on-site IT and ICT personnel may not Because edge sites typically do not have dedicated
be present. When reliable infrastructure design on-site IT and ICT personnel, remote management is a cru-
reduces future network issues, that can be cial capability, and consistency in deployments favorably
considered a valuable foundational first step. impacts that ability. When every site is exactly the same,
remote troubleshooting becomes easier. A technician can
Security: Edge environments are not traditional
• talk local personnel, who may not be trained in IT/ICT,
IT spaces. Controlling physical access ensures that through a series of steps to rectify a problem. Also, system-
authorized personnel can access the systems, while wide upgrades can be made, even by untrained personnel,
the curious passerby and nefarious characters with a single clear set of instructions based on the
are deterred. common architecture and layout used in all instances.
Remote Management: Intelligent infrastructure

• Intelligent Solutions
solutions deliver crucial oversight for remote Intelligent solutions play a fundamental role in edge
installations. Remotely monitor, manage, and installations. They support remote monitoring and man-
control physical and environmental conditions, agement, along with security, environmental controls,
thereby limiting the need for on-site visits. and power delivery. Intelligence can take many forms,
including systems that monitor connections within the
The specific infrastructure solutions that form these enclosure, temperature monitoring and control, security
building blocks include: access, and power distribution. Smart power distribution
units (PDUs) support many of these functions, as well
Preconfigured Solutions as sensors, intelligent patch panels and patch cords, and
When organizations have multiple sites, whether they are even intelligent door handles.
across town or on the other side of the globe, it is benefi- One of the fundamental needs of a data center
cial to have consistent layout, architecture, and footprint is security in order to prevent unauthorized access to both
at each of those deployments. The benefits include simpli- data and equipment. In traditional data centers, security
fied ordering, speed and consistency of deployment, and is managed by door locks and control systems that limit
ease of remote management. That consistency is achieved access. With an EDC that is contained in an enclosure
by preconfiguring the enclosure prior to arrival on site. on the factory floor, within a trailer near a mining site
This could use enclosures that are purchased preconfig- or in the manager’s office in a retail store, that same level
ured from a vendor or configured at a central site before of security is not always possible. A variety of systems are
being deployed around the globe. available that control access, whether from enclosure
Remote monitoring and management are also enabled via
intelligent solutions at the edge. This function is considered
by some to be the most critical because of the lack of dedicated
IT and ICT personnel at most edge locations.
handles that can be opened remotely by an authorized it comes to enclosures. Standard data center cabinets
user with a keycard or code on-site, as well as sensors that will work in some highly protected indoor spaces, while
send an alert to indicate unauthorized access. enclosures that are rated for harsh environments may
Remote monitoring and management are also be mandated for industrial applications. A mining site
enabled via intelligent solutions at the edge. This func- or refinery may seem like an application that requires
tion is considered by some to be the most critical because rugged enclosures, but those sites may use standard
of the lack of dedicated IT and ICT personnel at most cabinets inside a rugged container that has environmental
edge locations. When the nearest technician is several controls installed. Also, one healthcare clinic may require
hours away, downtime and travel become costly. This risk a small enclosure that will fit under a desk in a back closet,
pushes some organizations away from an edge decision. while another has a dedicated space that allows a more
When temperature, moisture or security sensors can traditional cabinet or rack.
advise of real-time risks and intelligent solutions allow When deciding about the type of enclosure, common
an operator on the other side of the globe to make adjust- characteristics to keep in mind include basics (e.g., size
ments in real time, those risks become manageable. Simi- and load rating), more specialized factors such as NEMA
larly, intelligent connectivity simplifies remote manage- or UL ratings, and the presence of or ability to add
ment; it can advise of a bad connection so adjustments environmental controls and security features.
can be made. Selecting the right size of enclosure is especially
Consider this example: A national sporting goods important. Much like Goldilocks, organizations need
retailer sets up kiosks in various parts of its stores during to find the one that is just right. A structure that is too
different times of the year. As summer season begins, large can eat up valuable real estate and would most
kiosks are moved into the boating and fishing area of the likely require a larger investment than necessary. More-
stores. Historically, this required a handful of technicians over, too small may mean the enclosure will not house
to travel to each location to change the connectivity all of the needed equipment today or in the near future.
to enable the point of sale system to switch to that new Beyond space for network equipment, the enclosure will
location. With an edge solution in a back room or office, need enough space for PDUs and network connectivity.
complete with intelligent connectivity, a technician can Right-sized for the environment means an enclosure
talk an on-site store employee through the changes that fits in the available footprint and houses all of the
needed—by activating lights at ports that need to be dis- needed equipment.
connected or connected. Within a day or two, every loca- Harsh environments will require specialized
tion has been changed, resulting in zero travel expenses enclosures that protect again dust, debris, moisture,
and less labor cost. vibration, and extreme temperatures.
Enclosures Cabling and Connectivity

The edge environment dictates the enclosure needed Standards-based cabling and connectivity help provide
for the deployment. It is imperative to understand the assurance that systems will function as designed. Whether
installation space because one size does not fit all when using optical fiber and/or copper, both the cabling and
30 I ICT TODAY
connectivity are critical for delivering reliability and York, San Francisco, Chicago). Applications that were
uptime. This is another area where the environment can running in these facilities were not particularly latency
impact choice. If systems are being installed in a harsh sensitive, and thus the distance of the colocation facility
environment or one where electromagnetic interference from a company’s office was of little concern. The edge
is present, then rugged jackets or shielding may be needed computing trend is changing that view. Some applica-
on the cabling and connectivity components. tions are becoming more sensitive to latency; thus, the
closer a facility is located to the end user, the better
USE CASES AND the performance of that application. Edge colocation
COMMON ENVIRONMENTS is taking advantage of the trend to move colocation
Each data center application and environment has unique facilities into more Tier 2 and Tier 3 cities. While not
characteristics that demand solutions that address those compromising latency sensitivity in the process, this
characteristics, and edge deployments are no exception. is allowing companies with operations in more rural
The infrastructure for a highly protected indoor environ- locations to gain the benefits of colocation.
ment is vastly different than what is needed in a harsh
space. Following are common use cases where edge Health Care
is deployed, along with general information on what Patient-generated health data (PGHD) is an excellent
has been defined as the four environments for edge: healthcare use case. In a perfect world, data generated
by IoT technologies, such as wearables, blood glucose
Use Cases monitors, home scales, telehealth tools, mHealth apps,
and other sensor-related devices, could be collected and
Edge Colocation analyzed to pattern one’s health, allowing for detailed
The colocation market has been growing for decades. For preventative healthcare and event detection. Patient
the most part, growth has been driven by colocation data could be analyzed at a macro level to detect
facilities located in or near larger Tier 1 cities (e.g., New patterns in a multitude of different areas (Figure 3).
FIGURE 3: Areas, such as surgical spaces, need to find ways to incorporate advanced IoT and PGHD sensor technologies while
ensuring patient privacy and security in data collection.
In 1996, the Health Insurance Portability and
Accountability Act (HIPAA) was passed in the United
States. It addresses concerns about sensitivity of health-
care information and the lack of security in place for that
data. HIPAA restricts how healthcare information can be
It is imperative to understand
transferred and analyzed. The challenge is in transferring the installation space because
one size does not fit all when
PGHD securely within the boundaries that HIPAA necessi-
tates. Published reports show that an estimated 40 per-
cent of IoT sensors will be PGHD devices, so this is going it comes to enclosures.
to become an industry-wide challenge very quickly.1
Industrial/Manufacturing
The industrial automation space has been collecting and
analyzing data for decades. This data has historically been
processed in centralized data centers, which in some cases factory floor. Contamination from water, dust, and
can be in a location quite some distance from the factory corrosives become potential hazards. Harsh environment
floor. Many newer industrial automation applications enclosures, equipment, and physical infrastructure are
require real-time or near real-time interaction with com- required to properly address these issues (Figure 4).
pute resources. Due to this requirement, many factories
are moving the compute resources closer to the factory Retail
floor or on the factory floor. This opens up a whole other The brick and mortar retail space has been under assault
realm of issues when data center servers and networking from online retailers for quite some time. To counteract
equipment are exposed to the harsh environments of the this assault, brick and mortar retailers are always searching
FIGURE 4: Many industrial automation applications require real-time or near real-time interaction with compute resources.
32 I ICT TODAY
for advantages they can leverage to provide
a competitive edge. A lot of the innovations
driving this competitive edge require a signifi-
cant amount of additional compute power
at the retail location.
•
Augmented reality (AR) mirrors can
show shoppers how they look in differ-
ent clothing without physically trying
them on (Figure 5).
 Amazon has patented a mirror that
dresses viewers in virtual clothes.
 Magic Mirror is an independent free-

standing console with a digital screen
that allows people to try on different
outfits using simple hand gestures.
•
Coca-Cola Freestyle machines have
servers attached to them that collect
and process customer preferences
FIGURE 5: Augmented reality mirrors make it possible for shoppers
on site; it then sends information
to virtually “try on” clothing, thereby potentially improving revenue for
to the cloud. retailers. This technology demands more compute power at the location.
•
Amazon Go stores that are brick and
mortar enable customers to grab items
off the shelves and simply leave by using
technologies, such as computer vision,
sensor fusion and deep learning.2
Agriculture
Smart farms are becoming the standard in today’s
ultra-competitive agricultural landscape. Any
advantage that can be obtained by technology
is directly attributable to higher yields, thereby
resulting in higher profits. Many companies are
competing in this space to provide farmers with
the best and most comprehensive data about
their fields. This is allowing farmers to make
informed decisions about planting, irrigation,
and harvesting based on multiple data sources,
which include direct sensor data from the fields
and even satellite data (Figure 6).
FIGURE 6: Many companies are competing in the agriculture space,
where farmers are using comprehensive data about their fields to make
informed decisions about planning, irrigation, and harvesting.
Edge Environments General Indoor
Four types of physical infrastructure environments are The general indoor environment may also be somewhat
prevalent in edge compute deployments. These environ- familiar. These are typically located in indoor spaces that
ments have characteristics that set them apart from others require environmental controls; however, they may have
and help define the solutions that are most advantageous space limitations and have enhanced needs around secu-
for each. rity and remote management. Examples of this environ-
ment would be a telecom room, server room, or closet.
Highly Protected Indoor The solution and vertical industry examples include:
The highly protected indoor environment is very familiar •
Indoor deployment with controlled
to those who have worked in traditional IT spaces. It is environment but with limited cooling
basically the data center space that many organizations and basic security
have on-site and commonly referred to as an on-premise •
Example: Retail stores, healthcare facilities
data center. However, this may be located in an area not
typically used. This environment has most or all the ame- Harsh Indoor
nities of a traditional data center space, including power, A harsh indoor environment is typically found in manu-
cooling, connectivity, physical security, and a highly facturing and more industrial-type settings. A growing
protected setting that is expected in a traditional data number of applications on the manufacturing floor
center. What it may lack is the trained data center staff require lower latency to perform at an optimum level.
to handle every need, so remote management is key To meet this requirement, equipment is placed in closer
to a successful deployment. proximity to the end use. In this environment, the
An excellent example of this environment is edge enclosure and connectivity may need to be rated for
colocation. Several colocation companies have adopted a harsher environment to protect equipment and data
the edge as their business strategy. These colocation data center performance. The solution and vertical industry
centers are in non-traditional second and third tier cities, examples include:
closer to the end user/use. The solution and vertical •
Limited protection from the outdoor
industry example include: environment and limited security;
•
Remote data center or similar deployment, potential for dry or liquid hazards
including highly controlled environment (e.g., dust, water)
and room/data hall level security •
Example: Manufacturing floor,
•
Example: Edge colocation DC warehouse
DESIGN AN EXCEPTIONAL
INTELLIGENT BUILDING.
RELY ON AN EXCEPTIONALLY
SKILLED BICSI RCDD ®.
Visit bicsi.org/rcdd for details.
34 I ICT TODAY
Outdoor AUTHOR BIOGRAPHIES: Brian L. Kelly joined Panduit in
When compute equipment is required to be placed 2012. Prior to joining Panduit, he worked for 14 years in the ICT
in an outdoor environment, it must be either placed and data center colocation industries. Currently, Brian manages
in an enclosure that is able to provide a “data center” the Network Architecture Research team as part of Panduit’s
like space (e.g., container, trailer) and/or be hardened Corporate Research and Development department. The team
to the point that it can withstand the extremes that the develops reference architecture content for the Data Center,
outside environment can bring (Figure 7). When dealing Enterprise, and Industrial Automation businesses. To date, Brian
with the outdoor environment, several factors must be has written well over 20 technical papers. He is a member of the
considered in the selection of the equipment and Converged Plantwide Ethernet (CPwE) architecture team which,
enclosure including: along with Rockwell Automation and Cisco, publishes design and
•
Temperature control (heating, cooling, implementation guides for the industrial space. Brian can be
solar loading) reached at brian.kelly@panduit.com.
•
Sealing (against liquids, particulates, Mark Dehmlow is a Sr. business development manager with
corrosives, flora and fauna) Panduit’s Data Center Business. In this role, Mark is focused on
•
Physical security (access control,anti-vandalism) global strategic accounts, enterprise on-premise and edge
•
Electromagnetic shielding applications for data center infrastructure. Mark is closely
•
Vibration and shock isolation following the evolution of distributed compute network
•
Example: Agriculture, mining architectures and the focus of organizations to leverage Hybrid
IT models as a result. Mark has over 20 years of experience
PUTTING EDGE TO WORK in the ICT sector, having worked with network operators and
Edge compute is changing the way that network and in technology distribution while serving in various product
compute resources are deployed. Many new and unfamil- management, product, segment and channel marketing roles
iar decisions will have to be made to have a successful in high tech manufacturing. Mark is particularly interested in
edge compute deployment. Physical infrastructure assisting users of all types in the adoption of technology solutions
is foundational to edge deployments and can enable to drive intended business outcomes for their organizations. Mark
consistency and reliability across a geographically holds a BA in business management from Bethel University and
distributed edge footprint. an MBA in international business from European University,
Brussels, Belgium. He can be reached
at mark.dehmlow@panduit.com.
REFERENCES:
1. Jordan, Nick.” 7 Staggering
Stats on Healthcare IoT
Innovation,” Data Driven Investor,
16 August 2020.
2. Tillman, Maggie. “What is
Amazon Go, where is it, and how
does it work?” Pocket-lint,
25 February 2020.
FIGURE 7: Outdoor spaces, such as refineries, require a container of some type to house
data center equipment.
PROTECTING THE
FUTURE AT THE EDGE:
Addressing the Risks of Emerging
Edge Data Centers
By Jacques Fluet
With 75 billion connected devices expected by 2025 as forecasted by Statista, 5G deployments being rolled out
and the promise of a host of new technologies, the ICT industry will see an unprecedented amount of data that
requires rethinking the way data center infrastructure is built and managed. To meet the needs of new applications,
data needs to be hosted very close to users, and network functionality needs to be performed in milliseconds versus
having data processed and stored in centralized data centers located sometimes many cities away from its source.
36 I ICT TODAY
Latency, which is essentially the time it takes for data real-time processing of data. National Highway Traffic
to traverse a network to its destination, is measured in mil- Safety Administration (NHTSA) studies indicate the aver-
liseconds (ms). Due to the way data travels from switch age reaction time of a human in traffic is around 1 to 2
to switch to server and back again, latency is unavoidable seconds. For self-driving vehicles to be safe and reliable
and varies based on the transmission medium, distance, and to gain consumer confidence, they require 5G’s 5 ms
number of switch hops and overall efficiency of the infra- latency, which is practically real time.
structure and equipment. While 50 to 200 ms to ping One part of the solution to reducing latency lies in
Google or 10 seconds to load a web page is something edge computing that brings IT and cloud-based services
most people are used to, there is no place for latency closer to end users and devices, which is why there is rapid
when it comes to supporting next-generation 5G, inter- growth in the deployment of edge data centers (EDCs)
net of things (IoT) and industrial internet of things (IIoT) closer to the source of the data being processed—any-
technologies like self-driving vehicles, smart traffic sys- where from the central office, 5G cell tower or local street
tems, virtual and augmented reality, artificial intelligence corner, to the enterprise office, manufacturing facility or
(AI), machine-to-machine (M2M) communication, remote mining operation (Figure 1). According to Gartner,
enhanced video streaming and online gaming. the global market for EDCs is projected to triple over the
A self-driving car needs to be able to respond virtually next five years with 75 percent of enterprise-generated
instantaneously to smart traffic signals and road sensors, data expected to be created and processed outside
and M2M devices rely on guaranteed response times and of traditional centralized data centers.
Cloud
Data Center
Data Processing Data Analytics Data Cashing

(Storage, Capture, Reporting) (Machine Learning, AI, M2M) (5G, E-commerce, Streaming)
Edge Data Center
Internet of Things
FIGURE 1: Edge data centers are located closer to users and devices to support emerging low-latency applications.
While smaller than a typical data center, EDCs vary Centralized Core or Cloud Sites: These network-
•
in size. They can be small micro-edge data centers with based sites typically perform local data collection
1 to 8 racks (or cabinets) and a power draw of 6 to 8 kW/ and processing, such as at central offices and
rack, medium with 4 to 12 racks with a power draw of 8 internet exchanges. Considered high-value edge
to 12 kW/rack or large with 12 to 49 racks and a power data centers, centralized core or cloud sites are
draw of 12 to 14 kW/rack. Regardless of size, the most typically large staffed facilities with higher avail-
common types of EDCs include the following network ability ratings and lower latency requirements.
and establishment-built sites:
Radio and Access Sites: These are small network-
• Figure 2 illustrates the relationship between latency
based EDCs with lower latency requirements, such and the network-based edge location size.
as those located in containerized purpose-built
shelters, cabinets or enclosures at a 5G cell tower, Establishment-Based Sites: These are typically
•
small cell location or other sites used for data small-to-medium size private-network driven
collection and immediate or short-term data EDCs that orchestrate transmission needs
processing. These sites are often unmanned between access and aggregation sites, while
and self-contained. performing local data collection and processing.
They are typically located on premises, serving
Aggregation Sites: These are medium-to-large
• enterprise businesses, industrial sites, large
network-based EDCs that are typically comprised retail establishments, financial sites, and
of multiple access networks and located at a point transportation hubs.
prior to traffic being routed to the regional data
center, such as meet-me-rooms for carrier hando- Given the critical nature of these spaces, special con-
ver points. Their primary purpose is user data col- sideration must be taken to protect EDCs and the equip-
lection, data aggregation, and analytic processing ment and components that reside within them. To that
and transmission. In some cases, aggregation sites end, the Telecommunications Industry Association (TIA)
may require the deployment of containerized or Edge Data Center Working Group is working to establish
purpose-built, prefabricated edge data centers. best practices for EDCs that include several of the
Radio Site Access Site Aggregation Core Site

Site
Increasing size of an edge
Lower latency
FIGURE 2: Network-based edge locations size and latency.
38 I ICT TODAY
following key considerations for addressing risk and to reliability or systems integrity should be considered
ensuring security and protection, from site selection and in this process.
physical security to thermal management and operations. Common environmental risks like severe weather,
seismic activity and natural disasters can significantly
SITE SELECTION CONSIDERATIONS affect operational continuity in even the most hardened
Data center site selection has always posed its own chal- of EDCs. For small, unmanned and/or only remotely
lenges with considerations that include both environ- monitored EDCs, it is more important to plan for and
mental and created outside influences, available utility mitigate the potential exposure to the elements.
services, and specific facility requirements. Because Aside from natural occurrences, unexpected man-
EDCs need to be located in a way that allows data to made risks outside the control of a site owner can also
be processed close to users and network functionality factor into site selection. Risks caused by adjacent uses,
to be performed in real time, site selection becomes even accidents, and potential acts of violence, theft or harm
more critical. should be considered when evaluating a potential site
First, it is important to understand that there is no location. In addition, the availability of skilled labor
such thing as the perfect site, and any location can have or support personnel capable of working to develop and
some element of risk. It is also important to understand maintain the EDC may also be a key factor to consider
what those risks are for any location being considered for in site location. Having a good understanding of the
an EDC, as well as the potential risk introduced to the potential risks, as well as understanding the potential
surrounding area as a result of its planned deployment. for use and population changes around a site, can help
It is highly recommended to ensure the right support in better plan for those potential risks. For example, nearby
the planning stages of the site selection by partnering hazardous conditions or manufacturing sites for hazard-
with consultants with strong experience in this area. ous products or materials could pose significant risk
The key to selecting a site for an EDC is to weigh both to the proposed site, as could adjacent travel corridors
environmental and manmade risks and decide if they are and proximity to civic, military or other socio-strategic
worthwhile investing in the site’s development. Loca- locations. See Table 1 for recommended distances from
tional criticality, business needs, costs, and implications adjacent facilities.
Recommended
Use/Facility Type
Distance
Airport (Approach/Take-off corridors) 5 Miles
Agriculture Uses (Grain Elevators, Livestock, Fertilizer/Machinery Storage) 5 Miles
Automotive Body or Repair/Machine Shops/Gas Stations/Truck Stops 1 Mile
Chemical Plants or Storage/Foundries/Conventional Power Plants 5 Miles
Embassies/Political Groups/Large Urban Centers 3 Miles
Landfills/Waste Storage or Treatment Plants 2 Miles
Military Installations or Munitions Storage 10 Miles
Nuclear Power Plants (Active) 50 Miles
Overflow/Run-off Areas for Reservoirs or Man-made Lakes 2 Miles
Quarries/Mines/Blast Zones 3 Miles
Radio/Television Transmitters/Stations 3 Miles
Research Laboratories 3 Miles
Self-Storage Facilities 1 Mile
TABLE 1: Recommended distances from edge data center sites to adjacent facilities.
The key to selecting a site for an EDC is to weigh both
environmental and manmade risks and decide if they
are worthwhile investing in the site’s development.
Adequate utility coverage is also an important factor, a way to lessen impact. For example, in areas like indus-
which can be a challenge for some remote locations. Con- trial, mining, coast, dumps, and oil rigs where equipment
nectivity is the primary need for every EDC site being could have potential risks of corrosion, explosion, and
considered, including the quality and bandwidth of that failures due to chemical agents in the air, air treatment
connectivity. While relatively low compared to more and corrosion protection solutions should be deployed
traditional data center sites, power needs for EDCs are to mitigate these risks. In areas of potentially high snow
also an important aspect of site developments. Water load, roofs may require a slope design, while high fire risk
availability can be a concern for EDC sites that require areas should consider fire-rated enclosures in addition to
water-based cooling systems. required internal fire detection and suppression systems.
Aside from the obvious need to get permitting for Physical protection measures also need to be deployed
an EDC site, other factors, such as local, state and federal to protect the EDC, as well as the components within and
regulations, may also impact a site’s development. Zoning any incoming service. Every data center should have
or planning guidelines for everything from parking and an access control system in place to ensure that gaining
vegetation to environmental permitting may be required. control to assets is authorized and restricted based on
Labor agreements, taxation, and safety requirements organizational requirements. Access control levels for
may all affect a site location’s ultimate viability. Having EDCs include:
a firm understanding of the regulatory requirements Level 1: Key-based with manually recorded
•
of a specific site is, therefore, critical for site selection. logon entry
PHYSICAL SECURITY CONSIDERATIONS Level 2: Electronic access with automated

•
There are several areas to consider when ensuring physical recording of entry/exit enabled by cloud-based
security of an EDC. While primary environmental risks third-party systems
should have been considered during the site selection,
it is virtually impossible for every location to be free of all Level 3: Electronic access, integrated with
•
environmental factors. Furthermore, unexpected condi- identity management system and automated
tions can arise, which can be anything from naturally recording of entry/exit with no third-party
occurring events like floods, solar flares and earthquakes cloud-based systems
to changes in the environment due to adjacent space
development. It could also be due to changes in the envi- Level 4: Electronic access, dual-factor authentica-
•
ronment that have required policy changes, such as recent tion (e.g., video verification, biometrics) with
COVID-19 policies that would cause a need for amending automated logging of entry/exit/location and
operation or mandating a maintenance, repair and no third-party cloud-based systems
upgrade necessary to sustain a safe and secure operation.
When considering potential environmental factors for Video surveillance has also become an important part
a given site, the EDC may need to be designed in such of EDC security and serves as the first layered application
40 I ICT TODAY
to support access control methodologies. A tiered to NIST/CIS framework automatically applied,
approach for video surveillance capability can also as well as yearly testing for vulnerabilities and
be used, including: a retention period of video for 180 days or longer
Level 1: IP-based video surveillance of outside/
•
perimeter with recorded video stored off-site It is also important to consider methods of
for limited retention period securing cable in pathways and at any ingress/egress
points, as well as any critical connection or splice
Level 2: IP-based video surveillance of outside/
• point located inside or outside of the EDC. The risk
perimeter with cybersecurity protocols manually assessment of vulnerability of incoming cabling
implemented and a retention period of video infrastructure should include not just the perimeter
stored off-site for longer than 30 days of the EDC but also the complete path of the cable,
including manholes, tunnels, maintenance holes, and
Level 3: IP-based video surveillance of both
• diverse distribution points along the path. Pathways
outside and inside of enclosure with cybersecurity for incoming data and power cables should prefera-
protocols to National Institute of Standards and bly be diverse and underground. For enclosure-based
Technology/Center for Internet Security (NIST/CIS) EDCs, the selection of the enclosures should even
framework automatically applied and a retention consider the material and construction in relation
period of video for 90 days or longer to tampering or unintentional damage. Depending
on location, an EDC may even need to be outfitted
Level 4: IP-based video surveillance of outside and
• for protection against animals and insects, which
inside of enclosure with cybersecurity protocols can include screens added to any ingress points.
Innovative Cable Pathway Telescoping Trapeze Supports J-Hooks, Trees and Assemblies
Connect and Protect your Datacom Project with nVent CADDY

CABLE PATHWAY, ROUTING AND HANGING SOLUTIONS FOR EVERY DATA & NETWORKING APPLICATION
nVent CADDY has a complete range of trusted, contractor-inspired solutions for your cable management needs. With the addition of our latest product line –
nVent CADDY WBT Performance Cable Tray – you can increase jobsite efficiency while working with our industry-leading, premium product portfolio.
Learn more about nVent CADDY datacom solutions: go.nVent.com/datacom
©2020 nVent. H86282-USEN-2008
The use of fences, trenches, berms or other manmade or may not be physically feasible. Therefore, thermal
or earthworks to create a perimeter or barrier around the management of an EDC must be considered for each
EDC site can further prevent access by unauthorized indi- level of cooling and heating, including:
viduals. Air conditioning condensing units, generators or Component Level: The component is the base
•
other auxiliary equipment should also be placed within of the heat source. It often relies on conduction
these barriers. Depending on location, no trespassing and to a case mounted heatsink and printed circuit
limited access signage is recommended to indicate private board (PCB).
property. Signage that highlights any potential risks from
exposures to electricity, light and radio waves and pro- Equipment (chassis, blade) Level: The equipment
•
vides contact information for site owners, administrators must be designed to effectively remove heat from
or management is also recommended. the component. It often relies on fans and air to
remove the heat directly from heatsinks and PCBs
THERMAL MANAGEMENT WITHIN THE EDGE and, therefore, indirectly from the components.
Heat that is externally present or generated internally
by the device and not adequately removed is detrimental Rack/Cabinet/Container Level: Plan equipment
•
to the performance and reliability of a communication layout to provide a thermal path for removing heat
and computational electronic circuit. That is why ther- from the equipment according to the thermal
mal management is one of the core subjects of any tradi- management design of the floor (e.g., equipment
tional data center infrastructure and critical to protecting stacking arrangement, heat baffle, thermal duct,
active equipment. fans, liquid bath).
Physically locating IT and cloud-based services closer
to end users and devices as described in the EDC concept Floor Level: Plan the floor to maximize the perfor-
•
means that a high-capacity, data center-like environment mance of thermal management while minimizing
will often be required in public locations, such as office the cost.
buildings, shopping centers, school campuses, event are-
nas and wireless cell sites. There are several considerations Overall Environment: Consideration of tempera-
•
when planning a cooling solution for an EDC that will ture and humidity in and around the environment
maximize compute, caching and networking performance, should also be considered.
including cost as it may not justify the benefit of an EDC
Considerations for the thermal management of data
storage and transport equipment includes the methods
and systems sufficient to maintain the equipment operat-
ing environment (temperature and humidity) within the
Server, storage, and switch physical operating condition range. Unless equipment has
requirements stating otherwise, EDC equipment with sup-
gear power levels are on the ply air should be controlled to meet ASHRAE TC9.9 guide-
rise, thereby driving higher lines, which include one recommended and four allow-
able ranges (A1 to A4). ASHRAE’s recommended range for
power dissipation in racks all equipment is 18°C to 27°C (64.4°F to 80.6°F). Allow-
and making cooling more able ranges A1 to A4 as compared to the recommended

range can enable potential cost savings for environments,
of a challenge. most notably those relying on the free cooling technolo-
gies. Typically, these environments will operate at upper
ASHRAE limits, reducing cooling need and expense.
42 I ICT TODAY
Allowable ranges, depending on equipment, extend from -40°C to +65°C (-40°F to +149°F) and Class 3 from
the temperature range to as wide as 5°C to 45°C (41°F -40°C to +70°C (-40°F to +158°F). Equipment categorized
to 113°F) for A4. with these temperature ratings is safe to operate without
A1: 15°C to 32°C (59°F to 89.5°F)
• cooling methods that require chilled gas or liquid. Any
A2: 10°C to 35°C (50°F to 95°F)
• outdoor equipment outside of this range is considered
A3: 5°C to 40°C (41°F to 104°F)
• non-temperature hardened equipment and requires active
A4: 5°C to 45°C (41°F to 113°F)
• cooling to maintain its inlet air condition to meet its spe-
cific operating temperature and humidity range. Equip-
For equipment deployed in the traditional data center ment deployed with a liquid or immersion cooling system
environment, ASHRAE TC9.9 recommends new network- must meet its hardware compatibility requirement.
ing equipment designs to be front-to-rear airflow rated Server, storage, and switch gear power levels are
to a minimum of ASHRAE Class A3 (40°C) and preferably on the rise, thereby driving higher power dissipation
ASHRAE Class A4 (45°C). For networking equipment in racks and making cooling more of a challenge. There
where the chassis does not span the full depth of the rack, are differing views as to what the average power levels per
it is recommended to have an airflow duct that extends rack are today and will be in the future. This complicates
to the front face of the rack. Furthermore, ASHRAE TC9.9 things further because ICT professionals do not know all
recommends the equipment be designed to withstand the use cases planned in the new 5G network, which will
a higher inlet air temperature than the data center cooling drive processing power, the amount of storage and the
supply air; this applies if the equipment is installed in associated switches that connect these devices to the
an enclosed space that does not have direct access to the network fabric.
data center air cooling stream or the equipment has a side- If rack power levels will be in the range of 10 to 20
to-side air flow configuration inside an enclosed cabinet. kW per rack for the next five or even ten years, EDCs
Temperature hardened equipment is critical for the can likely employ more traditional cooling methods and
outdoor application with limited cooling capacity. Telcor- take advantage of an established supply chain that affords
dia GR-3108 General Requirements for Network Equipment scale deployments and well-understood cost models for
in the Outside Plant classify allowable ranges as Class 2 not only the cooling equipment, but also for the servers,
storage devices and switch gears.
However, if rack power densities
Integrated Network surpass the 20 kW level in the near
Communications for future, then developing and deploy-

ing EDC infrastructure with the more
Educational Applications traditional cooling methods that only
support less than 20 kW per rack
would decrease performance, effi-
From classroom audio to campus-wide communication, FrontRow makes
ciency and flexibility. If rack densities
communication effortless in any learning environment. We offer superior
solutions that are simple and easy to manage. are greater than 20 kW per rack, the
industry should be driving technol-
sroom Aud mpus Com e Managem & Emerg
las Ca vic fe
ty e ogy partners and a supply chain
m
De
C
io
nc
Sa
en
un
to support the need for these newer

t
ica t ion
cooling methods and large-scale

om n
municatio
C
deployments.
Learn more at www.gofrontrow.com
OPERATIONS AND
ADMINISTRATION PLAY A ROLE
Proper operations and administrative practices are a key Automation considerations
component of protecting EDCs and should include pro-
cesses, tools and technologies that help monitor and
that could help physical
assess ongoing risks. Defined security policies and proce- security operational needs
dures should be in place for monitoring and responding
to authorized and unauthorized access, as well as for coor-
at an EDC include advanced
dinating with local utilities and agencies regarding emer- networked technologies, such
as motion/occupancy sensors,
gency response and law enforcement. Periodic inspections
of the site, perimeter, enclosure(s) and ingress point(s)
from normal wear or damage caused by trespassers, theft dry contact sensors, and
or vandalism are highly recommended.
These processes may also need to consider that EDCs biometric access control.
may have slower speed of reaction among operations staff
and may need to cover multiple locations, especially for
unmanned sites, which presents the case for predictive
analytics and service-oriented software enabled by the
very technologies these EDCs support such as AI. critical practices in place to protect EDCs, the emerging
Security and protective measures should, therefore, technologies they aim to support like self-driving vehicles,
be derived by sensors and software, such as integrated smart traffic systems, virtual and augmented reality, tele-
power monitoring systems, temperature and humidity medicine, drone delivery, and M2M communication will
sensors at the room and cabinet level, leak detectors, air struggle needlessly as they gain adoption.
quality sensors, and vibration sensors. Automation con-
siderations that could help physical security operational AUTHOR BIOGRAPHY: Jacques Fluet is the edge data center
needs at an EDC include advanced networked technolo- program director at the Telecommunications Industry Associa-
gies, such as motion/occupancy sensors, dry contact tion (TIA). He has over 30 years of telecommunications and ICT
sensors, and biometric access control. industry experience in wireless, optical, and wireline networks.
Automated infrastructure management (AIM) He has been involved in products and services development for
solutions are also an excellent tool for detecting and equipment vendors for most of his career—from the design phase
reporting, in real time, any unauthorized changes to the to large deployments. He has held various leadership positions
patch cord connectivity. Moreover, AIM systems can leading engineering teams for Nortel and Ericsson and has been
even be integrated with security cameras, where present, involved in projects around the world. Since joining TIA, he has
to activate and record during unauthorized events. been involved in 5G, IoT, smart buildings and EDCs with a special
interest in new applications from various vertical domains enabled
CONCLUSION by next generation networks. Jacques can be reached
While this article covers the importance of physical pro- at jfluet@tiaonline.org.
tection, it should also be noted that cybersecurity is a key
consideration for EDCs that should be addressed. As orga-
nizations, such as TIA, BICSI, and AFCOM, continue to
address best practices for EDCs through standards devel-
opment and educational curriculum, there will be more
much-needed recommendations for ensuring that EDCs
remain operational, secure and reliable. Without these
44 I ICT TODAY
How to Compete
in the New Normal?

As the economy slows, can you still make money? Do you have an aggressive
material cost to make a profit?
Thousands of ICC Elite InstallersTM have been making good money with ICC
connectivity products. Elite Installers proved successful through the last
recession and will continue to prosper during this slowdown. Curious? Give
ICC distributors a call.
Some high-priced distributors are not willing to carry ICC products because they cannot
make as much margin as the big brands. No worries, we have hundreds of distributors who
would love to have your business. We do not sell direct, only through distribution partners.
Find them on icc.com/distributor.
800.422.6191 888.792.7463 800.847.5629 800.309.2322 801.484.5238 800.238.0787

Ohio Maryland Texas California Utah Nebraska
icc.com cs@icc.com Premise Cables . Workstation Outlets . Patch Panels . RCM . Fiber Optics . Residential Enclosures
© 2020, ICC
Multimode Fiber Technology,
Latest Market Trends and Standards
By G. Mabud Choudhury and John Kamino, RCDD
Whether it is transmission speed, reach or capacity, move into the 500-meter (m) data center space. Also
enterprise and data center networks have undergone described are how the latest data center application areas,
a remarkable evolution over the past decade. However, including colocation, edge data centers (EDCs), and
one thing has remained constant: the continued relevance telecom/central office data centers, are providing new
of multimode optical fiber (MMF) links as the most and growing addressable markets for the cost and
cost-effective short-reach solutions on the market. power advantages of short-reach MMF links.
This article is designed in two parts to provide ICT Part two provides the most recent standards update
designers and installers with the latest knowledge about for MMF in the enterprise as a convenient best practices
MMF. In part one, the technology trends behind MMF’s and standards reference. Combined, these two parts
success story are explored. It explains how cloud and provide ICT designers and installers with a compre-
enterprise data centers continue to drive strong demand hensive over-view of MMF, equipping them to offer
for multimode solutions, even as transmission speeds their customers the best this technology has to offer
increase and singlemode optical fiber (SMF) transceivers now and into the future.
46 I ICT TODAY
PART 1 — MMF TECHNOLOGY, of traditional enterprise links are less than 100 m, while
MARKET, AND INDUSTRY TRENDS virtually all traditional enterprise optical data center
links are less than 300 m.
VCSEL Light Sources Drive MMF’s Growth In applications with speeds less than or equal to 10
Vertical cavity surface emitting lasers (VCSELs) are gigabits per second (Gb/s), twisted-pair copper solutions
the light source of choice for today’s MMF transmission have had remarkable success. However, at speeds greater
over advanced optical multimode OM3, OM4, and OM5 than and equal to 10 Gb/s, optical solutions including
optical fiber. The VCSELs offer several advantages over
1
traditional fiber cable/transceiver links and active optical
light-emitting diode (LED) and edge-emitting laser (EEL) cables (AOCs) are typically used. For distances less than
technology typically used in singlemode links, including: 100 to 150 m, MMF is the solution of choice, while SMF
•
Low power is typically used for longer links. Direct-attach copper
•
Cost (DAC) cables are often used for reaches less than 5 m.
•
Power efficiency In 2018, IEEE 802.3 completed IEEE
•
Manufacturability Std 802.3cdTM-2018, which defined standards for
•
Integration 50GBASE-SR, 100GBASE-SR2 and 200GBASE-SR4
•
Reliability on 50 Gbaud lanes over OM3/OM4/OM5 operating
•
Testability at 850 nm.
•
Custom packaging In January 2020, IEEE 802.3 completed IEEE Std
802.3cm™-2020, defining two new standards for 400
Multimode optical fiber, with its larger core size Gb/s Ethernet over OM3/OM4/OM5 MMF. Moreover,
compared to SMF, can more easily capture light from 400GBASE-SR4.2 is notable, as it is the first multimode
a transceiver, thereby decreasing alignment costs. While solution that uses wavelength division multiplexing
the cost of SMF optical sources have declined recently (WDM) to increase the data rate traveling over a pair
due to advances in silicon photonics, the less stringent of fibers.
alignment costs for MMF relative to SMF extend to multi- Proprietary and multisource agreement
mode connectors. A small 8µm SMF core is also much (MSA) solutions often fill in the gaps for specific appli-
more susceptible to dust and contamination than a 50µm cations. Extended reach (up to 400 m for 40 Gb speeds)
MMF core, making MMF links more robust in enterprise of standards-based solutions are often available for
and data center networks. customers who may have the need to support longer
For distances of less than 150 m and up to 550 m distances. In 2017, the short wavelength division multi-
in some applications, the lower cost of connectivity, plexing (SWDM) MSA released specifications for 40 Gb
installation and ongoing maintenance, along with lower SWDM4 and 100 Gb SWDM4; commercial products
cost of 850 to 940 nanometer (nm) transceiver modules, became available at the end of that year. The 100 Gb
make the combination of laser-optimized MMF and BiDi solutions were introduced in the same time frame,
VCSEL-based transceivers the most cost-effective adding another 100 Gb/s MMF/VSCSEL solution
optical short-reach solution for enterprise and for duplex links.
data center applications.
On the Road to 400 Gb/s Ethernet
Ethernet Market and Fibre Channel Market Trends Currently, 100 Gb/s speeds are being used primarily
The Ethernet market continues to grow and move by cloud data centers and leading-edge enterprise data
to higher speeds. For traditional on-premises enterprise centers. A significant portion of these adopters are
data center and LAN applications, OM3/4/5 fiber and utilizing the 100GBASE-SR4 solution, underscoring
VCSEL-based optics dominate structured cabling and the continued cost/power/overall value proposition
point-to-point optical short-reach links. The vast majority of MMF-VCSEL solutions with increasing data rates.
100 GbE by Reach - Shipments
100-300 m 500 m 2 km 10-20 km 40 km
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023
FIGURE 1: Shipments of 100 GbE by reach, 2010-2023. Source: LightCounting Ethernet Forecast - September 2018.
Figure 1 shows LightCounting’s prediction for Furthermore, 400GBASE-SR4.2 is the first multi-
strong 100 GbE growth from 2018-2023 for both mode Ethernet standard that utilizes WDM and takes
MMF-based 100 to 300 m reach and SMF-based 2 km advantage of OM5 fiber’s wideband capability. In this
reach solutions. Both MMF and SMF solutions are solution, two wavelengths or lanes are used to transmit
expected to grow significantly. 100 Gb/s over each fiber pair. Four MMF pairs used
The predicted uptake of 100 GbE MMF modules, with two wavelengths/fiber leads to the new designa-
driven by the introduction of 100 Gb/s switch links and tion 400GBASE-SR4.2 where 400GBASE defines the 400
later by the development of 100 Gb/s servers, shows the Gb speed, SR denotes Short Reach, and the 4.2 describes
need for 400 GbE MMF modules. New Ethernet standards the number of fiber pairs, followed by the number of
are being developed to support this demand. wavelengths used. These modules have been introduced
400GBASE-SR16, an MMF-based 16-fiber pair in the market and are expected to be widely deployed
solution for 400 Gb/s applications has been standardized in enterprise and data center networks. SR4.2 operates
but is not expected to ship in large volumes. Two new over two wavelengths at 850 nm and 910 nm just like
400 Gb/s multimode solutions were standardized currently popular 40 Gb and 100 Gb BiDi modules.
in January 2020. In 2018, the IEEE P802.3cm 400 Gb/s Moreover, 400GBASE-SR8 operates at 850 nm over
over Multimode Fiber Task Force was created, tasked 8 pairs of MMF, each pair carrying 50 Gb/s at 850 nm.
with developing standards for 400GBASE-SR4.2 and The 400GBASE-SR8 transceivers are commercially
400GBASE-SR8. These solutions are based on 50 Gb/s available and have been adopted by some large
lanes on OM3, OM4, and OM5. hyperscale customers.
48 I ICT TODAY
Data Rate Ethernet Standard IEEE # Fiber Optical Reach (m)
# λ ’S
Gb/s Proprietary/MSA Module Standard/MSA/ Pairs Modulation OM3 OM4 OM5
Same
100 100GBASE-SR4 Standard 4 1 25G NRZ 70 100
as OM4
100 100G – SWDM4 MSA 1 4 25G NRZ 75 100 150
100 100G – BiDi Proprietary 1 2 50G PAM4 70 100 150
400 400GBASE-SR16 16 1 25G NRZ 70 100
50 50GBASE-SR 1 1 50G PAM4 70 100 Same
Standard
100 100GBASE-SR2 2 1 50G PAM4 70 100 as OM4
200 200GBASE-SR4 4 1 50G PAM4 70 100
Same
400 400GBASE-SR8 8 1
Standard 50G PAM4 70 100 as OM4
400 400GBASE-SR4.2 4 2 150
100 100GBASE-SR 1
IEEE 802.3 Same
200 200GBASE-SR2 2 1 100G PAM4 30? 50?
Study Group as OM4
400 400GBASE-SR4 4
800 800GBASE-SR8 Future 8 1 100G PAM4 50?
4/4/800/1600 4/800/1600GBASE-SRm.n. Technology 1/2/4/8 TBD 100G PAM4 100M OVER MMF
TABLE 1: Common high-speed Ethernet modules. Question marks=the supported link distance is still not determined.
Also in 2018, a 400 Gb BiDi MSA was formed. Prod- to satisfy the market needs of the largest enterprise data
ucts meeting this MSA are expected to also be compliant center customers.
with the 400GBASE-SR4.2 standard. Refer to Table 1 for On the storage side, Fibre Channel (FC) remains
common Ethernet modules. the dominant standard for enterprise networks. More
Figure 2 shows The Ethernet Alliance’s 2020 Ethernet than 90 percent of FC links utilize MMF and VCSELs.3
Roadmap,2 documenting the growth in Ethernet speeds The Dell’Oro Group4 estimates that 113 million ports
of FC switches and adapters have
shipped since 2001 and that 46.2
Ethernet Speeds million ports are still currently
10T in operation with continued
Highly Parallel growth of FC links; 16 Gigabit
Speeds
Fibre Channel (GFC) is being
Quad Speeds
Link Speed (b/s)
1T deployed at a faster rate than

400G Duo Speeds 8 GFC, and 32 GFC is beginning
Serial Speeds to ramp up.5
200G
The T11 Technical Committee
100G of the InterNational Committee
50G for Information Technology Stand-
25G ards (INCITS) recently completed
FC-PI-7 standard specifications
10G
for 64 GFC in small form-factor
2000 2010 2020 2030
pluggable (SFP) modules. It is
Standard Completed nearing completion of FC-PI-7P
Ethernet Speed Ethernet Speed Possible Future Speed specifications for parallel or quad
lane 256 GFC (4 x 64 GFC) and
FIGURE 2: The Ethernet Alliance’s 2020 Ethernet Roadmap. is currently developing FC-PI-8
specifications for 128 GFC serial lanes. Newly developed
Fibre Channel over Ethernet (FCoE) allows FC protocols For distances less than 100 to 150 m,
to run over Ethernet networks. Fibre Channel is com-
pleting development of a standard that maps FC
MMF is the solution of choice,
to Non-Volatile Memory Express over Fabrics (NVMe-oF), while SMF is typically used for longer
allowing increased network performance and lower
latency. Figure 3 shows the Fibre Channel Industry
links. Direct-attach copper (DAC)
Association’s (FCIA) Roadmap for Fibre Channel, which cables are often used for reaches
will continue to remain a major application area for
MMF and VCSELs.
less than 5 m.
1TFC
400GFCoE
512GFC Quad Lane
in QSFP
200GFCoE
256GFC
128GFC
100GFCoE 100GFCoE
64GFC
50GFCoE
40GFCoE
32GFC
25GFCoE
Serial Lane
16GFC in SFP
10GFCoE
8GFC
2005 2010 2015 2020 2025
Fibre Channel Speed Speed in Development Speed in Development FCoE Speed Speed in Development Possible Future Speed
FIGURE 3: Information from FCIA’s Fibre Channel Roadmap.
Data Center Trends 19% CGAR 2016-2021

The single biggest trend for data centers in the last
600 Traditional Data Center (-5% CGAR)
Installed Workloads in Millions
few years has been the accelerated migration to public

Cloud Data Center (22% CGAR) 6%
and private cloud data centers. Traditional non-cloud 500
enterprise data centers are becoming a much smaller
400
segment of the data center market; large enterprises 94%
are building their own cloud data centers, and smaller 300
enterprises are using cloud services. The widely cited 200
and well-regarded Cisco Global Cloud Index (GCI) 17%
2016-2021 predicts that by 2021:
6 100 83%
•
94 percent of workloads and compute instances 0
will be processed by cloud data centers, while 2016 2017 2018 2019 2020 2021
six percent will be processed by traditional FIGURE 4: Workloads/compute instance: traditional data center
enterprise data centers (Figure 4). versus cloud data center. Source: Cisco Global Cloud Index, 2016-2021.
50 I ICT TODAY
13% CGAR 2016-202

•
Worldwide cloud traffic will more than triple. The Cisco GCI uses the NIST definition of cloud
•
Hyperscale data centers will represent 19%
53 CGAR 2016-2021
computing in defining cloud data centers and defines
percent of all installed data center servers five essential characteristics of a cloud data center:
600
(Figure 5) and will Traditional Data Center
grow in number (-5% CGAR)
to 628, •
On-demand self-service
Installed Workloads in Millions
Cloud Data Center (22% CGAR)

up from 338
500at the end of 2016. 6%
•
Broad network access
•
Resource pooling
400
Networking in the Cloud 94%
•
Rapid elasticity
The term “cloud”300
can cover many definitions and • Measured service
meanings. The National Institute of Standards and
200
17% “The NIST Definition
Technology (NIST) developed Broadly speaking, cloud data centers support three
100 7 in
of Cloud Computing,” 83%2011, explaining that “cloud service models: software as a service (SaaS), platform
computing is a model
0 for enabling ubiquitous, con- as a service (PaaS), and infrastructure as a service (IaaS).
2016 access
venient, on-demand network 2017 to 2018
a shared2019
pool 2020 Cloud
2021data centers also support four deployment models:
of configurable computing resources (e.g., networks, private cloud, community cloud, public cloud and hybrid
servers, storage, applications, and services) that can cloud. Figure 6 provides a graphical depiction of a hybrid
be rapidly provisioned and released with minimal cloud model, which is increasingly relevant for the
management effort or service provider interaction.” enterprise market.8
13% CGAR 2016-2021
700 53% 60%

48%
600 44% 50%
Hyperscale Data Centers
500 38%
628 40% % Share
32%
400 of Data
27% 570 Center
30%
509 Servers
300
448 (Installed
386 20% Base)
200 338
100 10%
0 0%
2016 2017 2018 2019 2020 2021
FIGURE 5: Global hyperscale data center growth. Source: Cisco Global Cloud Index, 2016-2021.
Google
Multi Cloud
Azure
Private Hybrid Cloud AWS

Cloud
Enterprise Data Center Public Clouds

FIGURE 6: Model of a hybrid cloud.
The term “cloud” refers to the
way IT resources are configured and
Global Private vs. Public vs. Non-Cloud
Enterprise Application Workloads and Compute Instances
accessed, not the size of the facility.
A very wide range of data centers are
moving to cloud computing. The Public Cloud Data Center (32% CGAR)
450
Installed Workloads & Compute

accelerating growth of the cloud Private Cloud Data Center (11% CGAR)
can also be viewed through the 400 Non-Cloud Data Center (-5% CGAR)
Instances in Millions
prism of private clouds compared 350
to public clouds and non-cloud data 300
centers, as shown in Figure 7. 57%
250
The focus tends to be on public
200
cloud data centers since these data 32%
150
centers have the highest growth
100 35%
rate and tend to be operated by 46%
the largest technology companies. 50
22% 8%
However, private cloud data centers 0
2016 2017 2018 2019 2020 2021
are also projected to achieve double-
digit compound annual growth rates.
FIGURE 7: Private versus public versus non-cloud data centers. Source: Cisco Global
Private cloud data centers can Cloud Index, 2016-2021.
be on-premises or off-premises
(e.g., multitenant data centers, colocation data centers). by Google,12 most employ a mix of copper, short reach
Non-cloud data centers are the only part of the market (SR) multimode optics, and long reach (LR) or coarse
that is shrinking. wavelength division multiplexing (CWDM4/CLR4)
singlemode optics.13 The most critical metrics for
Hyperscale Data Centers intra-datacenter interconnects are cost, size, and
The Cisco GCI defines a hyperscale data center based power in order of importance.14
on meeting one of the revenue criteria: 9
Other valid definitions for “hyperscale” primarily
•
More than US $1 billion in annual revenue refer to the size of the physical facility and the long cable
from IaaS, PaaS or infrastructure hosting services runs required (e.g., reaches of more than 300 m and up
•
More than US $2 billion in annual revenue to 2 km). The term “mega data center” is also often used
from SaaS to define such installations.14
•
More than US $4 billion in annual revenue
from Internet, search and social networking Traditional Enterprise Data Centers
•
More than US $8 billion in annual revenue The rapid growth of cloud data centers and the decline
from e-commerce/payment processing of traditional data centers often leads to the misinter-
pretation that enterprise data centers are disappearing.
Twenty-four providers meet the Cisco GCI hyperscale In fact, a significant percentage of large enterprise data
criteria. These 24 hyperscale data center companies utilize centers are being transformed to cloud data centers. They
a wide variety of architectures and both MMF and SMF are utilizing server virtualization, containerization,
optical links. Some of the largest U.S. and China-based software defined networks (SDN), network functions
hyperscale cloud companies are currently deploying virtualization (NFV), analytics and automation to provide
MMF-based 100G-SR4 or 400G-SR8. Others have
10
cloud computing capabilities inside the company for
adopted parallel SMF 4-lane (PSM4) or a duplex SMF rapid on-demand self-service, broad network access,
CWDM solution. According to a 2017 OFC presentation
11
and measured service.
52 I ICT TODAY
Figure 8 shows the Dell’Oro Group’s five-year Data Center Ethernet Switch Revenue
forecast (2018-2022) for Ethernet switch revenue
15
$18
by data center segments. While the switch revenue Top 3 Chinese Cloud SPs
forecast from the overall enterprise market declines Telco SPs S Ps

.S . Cloud
Top 4 U
slightly, the forecast for large enterprises shows
significant growth. These large enterprise data $9 Rest of Cloud
centers are adopting private/hybrid cloud strate-
gies to develop the most cost-effective data center Large Enterprises
solutions based on their business needs.
Rest of Enterprise
Given the size/reach, architecture, data, net- $0
working and storage needs of the enterprise and
12
13
14
15
16
17
18
19
20
21
22
20
20
20
20
20
20
20
20
20
20
20
private cloud segment, MMF and VCSELs will
remain a cost-effective, low-power solution. FIGURE 8: Dell’Oro forecast on Ethernet switch for data centers.
Source: Dell'Oro Ethernet Switch Data Center 5 Year Forecast,
January 2018
Other Data Center Trends
•
Convergence of wireline and wireless services,
Telecom/Central Office Data Centers: Central offices such as converged access infrastructure,
are being transformed by a number of industry trends. 16
networking platforms, and applications.
These trends include: •
SDN and NFV, which are key technologies that
•
Increased bandwidth demand for next generation enable virtualization in data center and central
PON and 5G. office environments.
•
Edge computing – lower latency requirements Telecommunications service providers have different
driving compute needs closer to the network’s edge. initiatives for transforming their central offices to data
•
Network simplification for lower cost. centers. They are still in the early stages, but the market
potential is significant. The data centers being built
A major initiative in this area is central office inside central offices are optimally sized to take
re-architected as a datacenter (CORD). According to the advantage of the cost and power advantages
Open Networking Foundation, CORD hardware architec- of MMF/VCSEL solutions.
ture is based on “commodity servers interconnected by
a fabric of white-box switches; switching fabric in a spine- Edge Data Centers: The EDC market is growing
leaf topology for optimized east-west traffic; specialized rapidly, as the internet of things (IoT) adds an expo-
access hardware for connecting subscribers (residential, nentially growing number of devices needed to support
mobile and/or enterprise).” In other words, a cloud data local machine learning, analytics/AI, mobile data, and
center for subscribers (Figure 9). 17
applications requiring low latency (Figure 10).18
CORD Virtualized Central Office
Mobile Enterprise Residential
Cord Controller (XOS)
Subscribers
Mobile WAN
Access
Enterprise
Access Shared Server Resources ROADM (Core)
Residential
Access
FIGURE 9: CORD virtualized central office.
Global Edge Data Center Market 2018-2022
MARKET GROWTH KEY TREND KEY DRIVER

The global market is expected Increasing adoption of IoT. Increase in the use
to grow at a CGAR of over 19% of mobile data.
during the forecast period.
FIGURE 10: Global EDC market, 2018-2022. Source: Technavio.
54 I ICT TODAY
FIGURE 11: An illustration of edge computing.
While hyperscale cloud computing drives a central- data centers are another growing segment of the enter-
ized computing model, edge computing swings the prise cloud data center solution. Industry leaders use
pendulum back to distributed computing (Figure 11). 19
a wide variety of network connectivity, including MMF,
Edge computing currently uses a mix of copper, in their colocation data centers.
MMF, and SMF depending on size and choice of network
architecture. Future needs, changing architecture, and MMF Forecast is Bright
the dramatic increase of local data traffic from the IoT The continued relevance and growth of 100 GbE MMF
will likely require structured cabling and reaches that modules and new 400 GbE MMF solutions support the
continue to make MMF/VCSELs an effective option CRU market research projection of growth for MMF
in this rapidly growing market segment. shown in Figure 12.20
Singlemode optical fiber modules have become
Co-Location Data Centers: In addition to hyperscale increasingly relevant in the past few years, especially
and on-premises enterprise cloud data centers, co-location in the large hyperscale data center where link distance
Worldwide Multimode Fiber Demand Feb. 2018 CRU
2014 2015 2016 2017 2018 2019 2020 2021 2022
FIGURE 12: Worldwide MMF demand, 2014-2022.
MMF vs SMF - 25G to 400G Speed Modules Only
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
FIGURE 13: MMF vs. SMF 25 to 400 Gb/s modules, 2010-2021. Source: LightCounting.
and data rates increased and the cost structure declined. For example, the growth of OM3 fiber, now ubiqui-
Over the past five years, the percentage of MMF modules, tous in enterprise networks, was highly dependent
once the dominant optical module type, has declined on Ethernet standards like 10GBASE-SR and Fibre
relative to the percentage of SMF modules for data rates Channel standards like 4GFC and 8GFC. Demand
of 25 to 400 Gb/s. It is important to note, however, that growth for OM4 fiber is based on standards, including
the relative decline of MMF modules in comparison 10GBASE-SR, 40GBASE- SR4, 100GBASE-SR4 and
to SMF modules does not mean that MMF module ship- 16GFC/32GFC.
ments, themselves, are declining. Indeed, a LightCounting The work of these standards organizations, along
forecast (Figure 13) shows strong projected growth for with industry MSAs, is vital to meet market demands
MMF modules along with the growth of SMF modules. 21
for new applications.
The future for MMF links continues to be bright!
Evolving Ethernet Standards for Multimode Fiber
PART 2: STANDARDS UPDATE Often, new standards are developed or revised to take
FOR MMF IN THE ENTERPRISE advantage of higher performance capabilities of new
Enterprise network customers and vendors—in fact, the products. Within the last few years, for example, TIA
entire telecommunications/ICT industry—benefit from and IEC developed the OM5 specification for MMF. This
interoperable multivendor industry standards developed extends the 850 nm performance of OM4 fiber out
through the interaction and coordination between several to 953 nm in order to take advantage of evolving
important organizations. These include component SWDM multimode VCSEL technology.
and optical fiber standards from TIA and IEC, structured These standards bodies are also characterizing legacy
cabling standards from ANSI/TIA, ANSI/BICSI and OM3 and OM4 bandwidth to further support SWDM
ISO/IEC, as well as application standards developed applications in brownfield installations; for example,
by IEEE and Fibre Channel. IEC Subcommittee 86A, Fibres and Cables, published IEC
The ability to meet market demand for speed and 60793-2-10:2019 in May 2019. This document contains
transmission capacity standards cannot be overstated. specifications for all category A1 multimode fibers.
56 I ICT TODAY
EMB Guidance in IEC 60793-2-10:2019
5000
4500 OM5
4000 OM4
OM3
EMB (MHz-km)
3500
3000
2500
2000
1500
1000
500
0
840 850 860 870 880 890 900 910 920 930 940 950
Wavelength (nm)
FIGURE 14: EMB guidance for OM3/4 MMF between 850 and 953 nm according to IEC 60793-2-10:2019.
One of the major additions is effective modal bandwidth Nearing completion is T11 of FC-PI-7P, 256 GFC
(EMB) guidance for OM3/4 MMF at wavelengths between parallel (4 x 64 GFC) standard, which specifies 100 m
840 and 953 nm (Figure 14). reach for OM4/OM5 fiber and 70 m reach for OM3.
This was needed to support the work of the IEEE Letter ballot review has been completed and is pro-
P802.3cm 400 Gb/s over Multimode Fiber Task Force that ceeding to public review. Publication of the standard
specified the standard IEEE Std 802.3cm -2020. The stan-
TM
is expected before the end of 2020.
dard was approved by IEEE-SA Standards Board on January The latest Fibre Channel project is FC-PI-8/128GFC
30, 2020 and was published on March 30, 2020. (serial/single lane), which was initiated in December
Meanwhile, TIA’s TR42 Telecommunications 2017 and became an approved project in April 2018.
Cabling Systems Committee and its TR42.12 subcommittee, The FC-PI-8 MRD has a target technical stability
Optical Fibers and Cables, has adapted IEC’s fiber specifica- date of 2021, and product availability is targeted
tions; this includes adapting IEC 60793-2 (General) for 2022. The requirements include backward
to ANSI/TIA-4920000-C and adapting IEC 60793-2-10 compatibility to 32 GFC/64 GFC and 100 m reach
(MMF family) to ANSI/TIA-492AAAF. The latter standard for OM5/OM4.
consolidated five TIA MMF documents (TIA-492AAAA, This project will leverage the work in the IEEE
AAAB, AAAC, AAAD, AAAE) into a single document. P802.3ck 100 Gb/s, 200 Gb/s, and 400 Gb/s Electrical
Interfaces Task Force for the electrical lane specifications
Recent MMF Work in Fibre Channel and FCIA and the work underway in IEEE P802.3db 100 Gb/s,
The newest Fibre Channel standard is FC-PI-7, 200 Gb/s, and 400 Gb/s Short Reach Fiber Task Force,
64GFC, which specifies 100 m reach for OM4/OM5 as well as potential work in OIF for 100 Gb/s optical
fiber and 70 m reach for OM3. In July 2018, the T11 lanes. Key challenges remain to be solved; these include
FC-PI-7/64GFC (serial) committee of INCITS completed the technical feasibility of 100 Gb/s VCSEL links meeting
its first letter ballot vote with unanimous approval and 100 m OM4/OM5 reach requirements, meeting back-
some comments. FC-PI-7/64GFC was published plane loss budgets, and handling host printed circuit
in November 2019. board (PCB) loss.
During the February 2019 Fibre Channel meeting, 100 m references were included in IEEE 802.3cd, 50 Gb/s,
different multimode solutions for 128 GFC were 100 Gb/s, and 200 Gb/s Ethernet for 50GBASE-SR,
discussed. Based on a straw poll, there was unanimous 100GBASE-SR2 and 200GBASE-SR4. The supported
support for consideration of the following: link distance for OM5 is equal to OM4 because all
•
A true serial 128 GFC variant these links operate only at 850 nm.
•
A multi-wavelength variant 56 Gb/s
BiDi solution IEEE P802.3cm 400 Gb/s over Multimode
•
A multi-wavelength variant 56 Gb/s Fiber Task Force
codirectional (CoDi) solution The IEEE P802.3cm Task Force met in May 2018
to begin work to “Define a physical layer specification
One of the most interesting contributions made that supports 400 Gb/s operation over 8 pairs of MMF
in July 2019 was a technical proposal by Finisar, which with lengths up to at least 100 m.” Additionally, the
suggested that serial 128 GFC could be accomplished Task Force is to “Define a physical layer specification
by using VCSELs centered around 940 nm, rather than that supports 400 Gb/s operation over 4 pairs of MMF
the traditional 850 nm. with lengths up to at least 100 m.”
Meeting the target completion dates and all require- This was based on the work of the IEEE 802.3
ments per FC-PI-8 MRD will be very challenging; there Next-generation 200 Gb/s and 400 Gb/s MMF Study
are complex technical and market questions that will Group, which established the need for next generation
need to be resolved. multimode 400 Gb/s solutions. These solutions
were designed to leverage recent technology
MMF Developments in IEEE advancements, including:
•
Multiple wavelengths on MMF, first commercially
802.3bs and 802.3cd introduced in 2013 in 40 Gb BiDi modules
In December 2017, references specifying 100 m • VCSELs supporting 50 Gb/s 4-level pulse amplitude
reach for 400GBASE-SR16 over OM4/OM5 MMF modulation (PAM4) signaling, while leveraging
were included in IEEE Std 802.3bs™-2017, 200 Gb/s the historical multimode/VCSEL link advantages
and 400 Gb/s Ethernet. In December 2018, the same that include:
 Relaxed alignment tolerances
Purchase the 14th Edition TDMM (~10x) for lasers, multiplexers/

demultiplexers, and connectors
Detailing the Standards and Best Practices
 Greater resiliency to dirt
Related to Modern Cabling Design
 Lower drive currents
(5-10 milliamps [mA]
vs. 50-60 mA)
 On-wafer testing
VISIT BICSI.ORG/TDMM FOR DETAILS.
58 I ICT TODAY
Telecommunications service providers have different initiatives
for transforming their central offices to data centers. They are still
in the early stages, but the market potential is significant.
The study group found there was considerable 400GBASE-SR4.2 meets the four-pair objective
support for two different 400 Gb/s solutions.22
and uses two wavelengths per fiber, transmitting
1.
400 Gb/s over 4 MMF pairs up to at least bidirectionally (BiDi) on each fiber. Furthermore,
100 m, supported by an industry expert 400GBASE-SR8 meets the eight-pair objective and
affiliated with one of the largest Chinese does not use WDM, thereby requiring double the
hyperscale cloud companies number of fibers relative to 400G-BASE-SR4.2. Both
2.
400 Gb/s over 8 pairs of MMF, supported by use the 400GAUI-8 electrical interface and convert
an industry expert affiliated with one of the 8x50G electrical to 8x50G optical.
largest U.S. hyperscale cloud companies As mentioned earlier, 400GBASE-SR4.2 is the first
standard to use WDM in an MMF/VCSEL-based solution,
In May 2018, the 400GBASE-SR8 baseline was and it takes advantage of OM5 fiber’s higher bandwidth
adopted; two months later, the same group adopted the at longer wavelengths. Link distances of 70 m for OM3
400GBASE-SR4.2 baseline. Later, IEEE 802.3cm was
23
fiber, 100 m for OM4 and 150 m for OM5 are supported,
approved in January 2020. and a standard 12-fiber MPO connector with eight active
The two 400 Gb/s MMF physical medium dependent fibers is defined as the interface. It operates at two
sublayers (PMDs) are shown in Figure 15. Undoubtedly, wavelengths, 850 and 910 nm.
400G MMF Under Consideration in Ethernet 802.3cm

400GBASE-SRx.y
# fibers # wavelengths per fiber
400GAUI-8
400GBASE-SR4.2
electrical
Fewer fibers using WDM Both are 8x50G

electrical to
8x50G optical
400GAUI-8 400GBASE-
electrical SR8
No WDM but 16 total fibers
FIGURE 15: 400G MMF, standardized in IEEE 802.3cm.
Providing a low-cost point-to-point link, 400G-SR4.2 will also support a 4x100G breakout, although
100G-SR1.2 is not an IEEE standard. A 4x4 fiber shuffle allows a 32-port 400G switch to be used as a 128-port
100G switch (Figure 16).
Additionally, 400G-SR8 is an 8 fiber-pair solution supporting link distances up to 70 m on OM3 fiber
and 100 m using OM4/OM5. Either single-row MPO-16 or 24-fiber MPO (MPO-12 two-row) can be used as the
connector interface (Figure 17).
SR4.2 SR4.2 [LEFT] SR4.2 Low Cost Point-to-Point Link
SR1.2
SR1.2
[RIGHT] SR4.2 4 x 100G breakout SR4.2
SR1.2
SR1.2
SR4.2 SR4.2
[LEFT] 4 x 4 fiber shuffle allows

SR4.2 SR4.2 a 32-port 400G switch to be used
as a 128-port 100G switch
SR4.2 SR4.2
SR4.2 SR4.2
FIGURE 16: 400GBASE-SR4.2 point-to-point, breakout and shuffle modes.
FIGURE 17: 400G-SR8 adopted connector options for MDI.24
60 I ICT TODAY
Attributes of 400GBASE-SR8 optics offer maximum flexibility for fanout and shuffle applications. Figures 18
and 19 show point-to-point, 2x2 fiber shuffle, 4x4 fiber shuffle and 8x50 G breakout capabilities and applications.
SR8 SR8 [LEFT] SR8 Point-to-Point Link
[RIGHT] 2 x 2 fiber shuffle allows a SR8 SR8

32-port 400G switch to be used as 64
port 200G switch
SR8 SR8
SR8 SR8
SR8 SR8 [LEFT] 4 x 4 fiber shuffle allows

a 32-port 400G switch to be used
as a 128-port 100G switch
SR8 SR8
SR8 SR8
FIGURE 18: 400GBASE-SR8 point-to-point and shuffle modes.
SR
SR8 SR8
SR
[ABOVE] SR8 module shown in point-to-point link
SR
SR
SR8
SR
SR
[RIGHT] SR8 module shown in 8 x 50G breakout
mode to connect a 400G port on a TOR switch to 8 x 50G server ports SR
May be implemented over a few meters in AOC

SR
FIGURE 19: 400GBASE-SR8 breakout.
The key value proposition of SWDM4 and 100 Gb BiDi is the utilization
of much lower cost, often already installed, duplex fiber cabling
and connectivity instead of parallel fiber cabling and connectivity.
IEEE P802.3db 100 Gb/s, 200 Gb/s, yet specified), offering eight-way breakout from one
and 400 Gb/s Short Reach Fiber Task Force 800 Gb/s switch port (2 x 400 Gb/s) to eight 100 Gb/s
A task force has been formed to develop physical layer server ports.
specifications that will support a number of objectives, There is also considerable interest on the part of some
including 100 Gb/s over 1 pair of MMF, 200 Gb/s over hyperscale customers for switch-to-switch links up to 50+
2 pairs of MMF, and 400 Gb/s over 4 pairs of MMF, with meters. As manufacturing experience is gained, support
lengths up to at least 50 m. The key area of consideration for longer reach 100 m switch-to-switch links will
will be MMF/VCSEL-based solutions. become available.
Large hyperscale and enterprise data centers The market need for 100Gb/s lanes is expected
are considering a move away from top of rack (ToR) to expand rapidly once 100 Gb/s server ports become
switching, migrating to middle of row (MoR) or end available and are rapidly deployed in hyperscale data
of row (EoR) switches as server counts in racks decrease, centers. These data centers can contain hundreds of
driven by increasing power dissipation per server as server thousands of servers, so the availability of a low power,
speeds move to 100 Gb (Figure 20). Links over MMF will
25
low cost solution is crucial. When this requirement for
evolve to 100 Gb/s PAM4 VCSELs in the 2021 timeframe. higher speed is combined with longer switch to server
They will initially target short-reach (30 m) server links, traditional copper twinax and twisted-pair solutions
interconnects in an 800GBASE-SR8-type application are no longer viable, thereby opening the door for new
(2 x 400G-BASE-SR4, since 800 Gb/s MAC rate is not low-cost optical solutions.
Breakout Use Case - Servers

Today Future
TOR Tier 0/1

Switch Switch
ASIC ASIC
25G 100G
25G 20- 100G
25G 30m 100G
100G
3m 25G
Server Server
50G (2x25G)
50G (2x25G)
No Supporting SMF PMDs
Server
Consortium/MSA
FIGURE 20: Evolution from TOR switch architecture to MOR switch architecture.
62 I ICT TODAY
Potential Next Generation MMF Links BiDi products became available at the beginning of 2018
Current MMF technology allows speeds of up to 200 Gb/s and have since been widely adopted.
in a single fiber pair by combining WDM and advanced The key value proposition of SWDM4 and 100 Gb
signal modulation. As 100 Gb/s lanes become available, BiDi is the utilization of much lower cost, often already
four wavelength WDM can provide the capability to sup- installed, duplex fiber cabling and connectivity instead
port 400 Gb/s links over a pair of fibers and up to 1.6 Tb/s of parallel fiber cabling and connectivity.
over 4 pairs of fiber. Of course, significant work remains An MSA on 400 Gb BiDi was announced in July 2018
in order to support four wavelength, 100 Gb/lane/s WDM, to define optical data link specifications and promote the
primarily on the transceiver side. adoption of interoperable 400 Gb/s optical transceivers
based on a dual-wavelength bidirectional transmission
MSAs Play an Important Role technology over MMF. These parts are expected to be fully
In March 2017, an SWDM MSA was developed compliant and backward compatible with IEEE 802.3cm
and released with specifications for 40 Gb SWDM4 and 400GBASE-SR4.2 standards-based products.
100 Gb SWDM4. Commercial 40 Gb SWDM4 and 100 Gb The supported reaches for OM3, OM4 and OM5 fiber
SWDM4 products are currently available. These SWDM4 for 100 Gb-SWDM4, 100 Gb-BiDi, 400 Gb-BiDi, along
modules utilize four wavelengths with 30 nm spacing with MMF PMDs from 802.3bs, 802.3cd and 802.3cm are
(850 nm, 880 nm, 910 nm and 940 nm). shown in Table 2. Highlighted is the extended reach
Essentially, 100 Gb BiDi uses two wavelengths benefit and future-proofing value proposition of OM5.
(857 and 908 nm) to extend the value of duplex cabling It also points out the potential of 100Gb/s lanes for
and connectivity infrastructure. Commercial 100 Gb an 800 Gb solution.
Data Rate Ethernet Standard IEEE # Fiber Optical Reach (m)

# λ ’S
Gb/s Proprietary/MSA Module Standard/MSA/ Pairs Modulation OM3 OM4 OM5
Same
100 100GBASE-SR4 Standard 4 1 25G NRZ 70 100
as OM4
100 100G – SWDM4 MSA 1 4 25G NRZ 75 100 150
100 100G – BiDi Proprietary 1 2 50G PAM4 70 100 150
400 400GBASE-SR16 16 1 25G NRZ 70 100
≤30-50m
50 50GBASE-SR 1 1 50G PAM4 70 100
breakout
Same
Standard
as OM4 to server
100 100GBASE-SR2 2 1 50G PAM4 70 100
will be first
200 200GBASE-SR4 4 1 50G PAM4 70 100
Same
use of
400 400GBASE-SR8 8 1 100G/lane
Standard 50G PAM4 70 100 as OM4
400 400GBASE-SR4.2 4 2 150 VCSELs
100 100GBASE-SR 1
IEEE 802.3 Same
200 200GBASE-SR2 2 1 100G PAM4 30? 50?
Study Group as OM4
400 400GBASE-SR4 4
800 800GBASE-SR8 Future 8 1 100G PAM4 50?
4/4/800/1600 4/800/1600GBASE-SRm.n. Technology 1/2/4/8 TBD 100G PAM4 100M OVER MMF
TABLE 2: Evolution of VCSEL-based links over MMF. Question marks=still not determined.
CONCLUSION
Ultimately, successful technology will leverage both improved MMF and VCSEL-based transceiver properties in order
to minimize total link costs in the cost-sensitive short-reach enterprise/data center environment, preserving the
VCSEL-MMF link advantages far into the future.
The strong projected growth trends for MMF, in conjunction with standards advancements, ensure that MMF
solutions are here to stay—giving ICT designers and installers continually more flexibility for their customers’
cost-effective data center and enterprise high-speed applications.
AUTHOR BIOGRAPHIES: G. Mabud Choudhury is standards 10. C. Xie, Alibaba OIF Q4 2017 Shanghai
manager for OFS. He represents OFS in IEEE 802.3 Ethernet, 11. C. Gang, Baidu, 2018 Optinet, Shanghai
INCITS T11 Fibre Channel and IEC standards. Previously, he was 12. X. Zhou, Google, OFC 2018, San Diego
an R&D engineer and manager for product development, systems 13. Zuowei Shen, Google “400G-SR8 Broad Applications
engineering and copper standards, fiber and intelligent cabling for Datacenters” IEEE 802.3 NGMMF SG, March 2018
infrastructure products/solutions for AT&T Bell Labs, Lucent, Avaya, http://www.ieee802.org/3/NGMMF/public/Mar18/shen_
and CommScope. He has an MS degree in mechanical engineering NGMMF_01a_mar18.pdf
from Massachusetts Institute of Technology and a BS degree 14. Ryohei Urata, Hong Liu, Xiang Zhou, and Amin Vahdat,
in mechanical engineering and in chemistry from Duke University. “Datacenter Interconnect and Networking: from Evolution
He holds 32 US patents and has authored several papers, webinars to Holistic Revolution” ryohei@google.com OFC 2017
and IEEE/Fibre Channel/TIA standards contributions. Mabud can 15. Dell’Oro Ethernet Switch Data Center 5 Forecast. January
be reached at mchoudhury@ofsoptics.com. 2018. From http://www.ieee802.org/3/NGMMF/public/
John Kamino, RCDD, is a senior manager of Product Mar18/kolesar_NGMMF_01b_mar18.pdf (page 15)
Management for OFS. A 27-year veteran of the company, 16. “The virtualized and converged central office/cable
he has also held positions in optical connectivity product headend” White Paper, CommScope, February 2018
management, offer management, product marketing, sales, 17. CORD (Central Office Re-architected as a Datacenter)
and engineering. He has a B.S. in Chemical Engineering from platform https://www.opennetworking.org/cord/
the University of Nebraska-Lincoln and an MBA from Mercer 18. “Top Factors Driving the Global Edge Data Center
University. John can be reached at jkamino@ofsoptics.com. Market – Technavio” Business Wire, March 2018
19. “Future Ready: Top Data Analytics Platforms to Leverage
REFERENCES: Edge Computing” TechGenix, September 2017
1. Finisar. http://myvcsel.com/key-advantages-of-vcsel- 20. Worldwide Multimode Fiber Demand, CRU, February 2018
technology/ 21. LightCounting Market Research, August 2018
2. Ethernet Alliance. The 2018 Ethernet Roadmap. 22. Contributed Presentations for IEEE 802.3 Next-genera-
https://ethernetalliance.org/technology/2020-roadmap/ tion 200 Gb/s and 400 Gb/s MMF PHYs Study Group
3. FCIA, The Fibre Channel Roadmap January 22-23, 2018, IEEE 802.3 Interim Meeting,
https://fibrechannel.org/roadmap/ Geneva, Switzerland http://www.ieee802.org/3/NGMMF/
4. Dell’Oro, May 2017 public/Jan18/
5. FCIA, Fibre Channel Solution Guide 2017 23. Contributed Presentations for IEEE 802.3 Next-genera-
http://fibrechannel.org/wpcontent/uploads/2017/07/FCIA_ tion 200 Gb/s and 400 Gb/s MMF PHYs Study Group
SolutionsGuide2017_WEB.pdf March 6-7 2018, IEEE 802.3 Plenary Meeting,
6. Cisco VNI Global Fixed and Mobile Internet Traffic Roseland, IL, US
Forecasts. 7th annual Cisco Global Cloud Index (GCI) 24. Paul Kolesar, CommScope “400GBASE-SR8 MDI
forecast. Global data center and cloud computing trends Choices” IEEE P802.3cm TF, May 2018
(2016 – 2021). February 2017 http://www.ieee802.org/3/cm/public/May18/kole-
7. National Institute of Standards and Technology sar_3cm_01_0518.pdf
“The NIST Definition of Cloud” Special Publication 25. Tom Issenhuth and Brad Booth, Microsoft “100G-DR
800-145. September, 2011 Use Cases & End User Perspective” IEEE P802.3cd TF,
8. Cisco blog “AWS or Private Cloud or both, what’s your September 2016 http://www.ieee802.org/3/cd/public/
strategy?” November 2016 https://blogs.cisco.com/datacen- Sept16/issenhuth_
ter/aws-or-private-cloud-or-both-whats-your-strategy
9. “Cisco Global Cloud Index: Forecast and Methodology,
2016–2021 White Paper” Document ID:
1513879861264127. February 2018
64 I ICT TODAY
REGISTER TODAY!
IoT • 5G Networks • PoE Applications and Standards • Data Centres

AV over IP • Wireless Technologies • Artiﬁcial Intelligence (AI) Smart Cities,
Buildings and Infrastructure • Single-pair Ethernet • And More
bicsi.org/emea2020
CONTACT US: emea2020@bicsi.org PHONE: +971 4 516 3018
THE IMPACT OF ICT
ON HEALTHCARE SIMULATION
By Patti Fisher, RCDD, DCDC, OSP, RTPM, WD
Healthcare simulation utilizes artificial means to simulate clinical and procedural scenarios and provides medical
professionals and students seeking to become medical professionals with the means to practice skills and receive
feedback. Healthcare simulation not only educates but also helps the development of new medical devices and
processes. It prepares surgeons and medical teams for complex procedures and improves safety practices and
patient outcomes. Anyone who participates in a medical simulation is called a “learner.” Learners are often
students at a college or university, but they may be medical professionals seeking to improve or practice
a specific skill in a controlled environment.
66 I ICT TODAY
Simulation devices run the range of technology from low
fidelity (realism) to high fidelity (Figure 1). Readers may
have used a medical simulator if they have used Resusci
Anne to learn mouth-to-mouth resuscitation and chest
compressions in a first aid class. This low fidelity simula-
tion device has been used since the 1960s as a tool to train
medical professionals and is still used today.
Simulation can involve simple task trainers, artificial
high fidelity manikins or even humans (standardized
patients). Simulators have been developed for training
procedures ranging from the basics, such as blood draw,
to the more complex including surgery, maternal, infant
and trauma care. Simulations can focus on skills needed
by those hoping to work in the dental or eye care field
or any number of specialties (Figure 2).
In nursing schools, simulating entire hospital floors
is becoming more common—from patient rooms to nurse
stations, pharmacies to emergency rooms. Simulating
a whole hospital floor demands the need for additional
spaces, such as the ambulance entrance and additional
code blue and nurse call systems. The addition of other
simulated systems impacts the design of the control
room as well.
FIGURE 1: Nursing students with high fidelity
simulation manikin.
FIGURE 2: Dental simulation lab.
High fidelity simulators usually involve two comput-
ers connected to a life-size manikin that provides varying
biological-like responses to medical treatments. These
manikins recreate specific medical conditions, often with
intervention from instructors located in a control room;
the operator speaks through the manikin to provide
needed information for the experience (Figure 3).
Surgical trainers and some task trainers also connect
to computers.
Standardized patient simulations use human actors
that have been prepared to act out a specific set of symp-
toms. The actors are made up to exhibit the condition
they will portray. They typically have watched a video
prior to arriving at the simulation center. Before begin-
ning the simulation, they participate in an orientation
meeting which generally occurs in a patient lounge.
Video capability is usually required here; it is also com-
mon to have informational displays or kiosks to aid in
providing scheduling and location information in this
area (Figure 4).
Since the actors must remain isolated from the
learners, standardized patient lounges generally have
dedicated entrances from the building exterior, separate
corridors (monitored by cameras and having announce-
ment capabilities), separate locker rooms, restroom facili-
FIGURE 3: Infant simulation lab. ties and even separate doors into the examination rooms.
Isolation of the patients from the students begins at the
entrance to the building which remains controlled, often
allowing entry through the use of a video intercom.
FIGURE 4: Standardized patient suite.
68 I ICT TODAY
Students rotate through each simulation and after each
encounter, both the student and the patient complete
documentation typically on computers. The student
commonly uses an annotation station in the corridor
while the patient does their annotation on the in-room
computer (Figure 5).
Most large simulation centers have dedicated
control or observation rooms. These may be located
between every few simulation rooms or may consist
of a single large control room with multiple control
stations. Facilitators in the control room adjust devices
in the simulation space to mimic different medical
and environmental conditions. They provide com-
munication into the simulation rooms, observe the
learner’s performance and collect data (Figure 6).
Simulations can be held in a classroom, flexible
environment or in spaces specifically designed for FIGURE 5: Learner entrance to standardized patient exam rooms.
simulation practice. While some locations represent
a specific medical environment, such as a labor and of Kansas Medical Center, “when we have practicing
delivery room or operating room, others allow for learners participate in simulation events we attempt
enactment of a variety of events and medical needs. to have the equipment they use match what they have
According to Amy Follmer, certified healthcare simu- available in the real world. The more realistic you can
lation operations specialist (CHSOS) with the Zamierowski make a simulation room look for learners, the easier
Institute for Experiential Learning, a partnership of the it is for them to actively participate in the simulation
University of Kansas Health System and the University and buy into the learning process.”
FIGURE 6: Simulation control room/observation room.
When an operating room is simulated, the space will
be fitted out like an actual operating room with overhead
booms, clocks and elapsed timers. Overhead service col-
umns might be used to provide data to the center of the
room. For an ICU simulation space, headwalls are often
used to provide simulated services, such as vacuum and
medical gases. Often, simulated monitors are located
on or adjacent to the headwall.
In a virtual anatomy lab, 3D dissection can be
performed using a specialized table. The table provides
detailed views inside a variety of cadavers. This technol-
ogy provides a full-size view inside of a human or ani-
mal with touch screen capabilities, allowing the learner
to make cuts and view systems and organs without the
need for cadavers (Figure 7).
Basic components of a simulation system include
a firewall, storage, and capture cameras. These cameras
may be the same model as those used for security, and
each simulation space generally has multiple cameras
to capture the high level of detail from more than one
angle. Bidirectional audio is also captured, and there
is commonly a device to capture the high resolution
video as well.
In addition to the simulation spaces, control rooms,
FIGURE 7: Students learning through a virtual dissection table. and AV equipment rooms, simulation centers typically
include classrooms, annotation stations, debrief loca-
tions, manikin preparation and storage rooms. Class-
rooms introduce the skills to be simulated and, therefore,
should be adjacent to the simulation rooms. Annotation
stations are often located just outside of the simulation
rooms. It is here that students review their simulated
patients’ medical charts before beginning care.
To say that debrief locations are small group rooms
oversimplifies the purpose and needs of the spaces. “The
debrief after a simulation is the opportunity to review the
simulation experience and provide feedback in a positive
learner focused environment. Simulation debriefing
doesn’t just reveal what was done incorrectly, it exam-
ines the clinical judgment and critical thinking behind
what happened in a way that supports learning,” says
Ms. Follmer. “You can watch a simulation and identify
that a learner gave the wrong medication, but until you
discuss with them why it happened, it’s hard to transfer
FIGURE 8: A debrief room. the learning to future situations.” (Figure 8).
70 I ICT TODAY
It is important to ensure the debrief spaces have suffi-
cient technology, including high quality video and audio In addition to the simulation spaces,
capabilities, lighting, acoustical management, and video
conferencing capabilities so personnel who are unable control rooms, and AV equipment
to be in the room can participate in the debrief. rooms, simulation centers typically
Space is required for manikin preparation and storage.
This space typically needs network connectivity so that include classrooms, annotation
the manikins can be updated. Often sufficient counter
stations, debrief locations, manikin
space is installed to provide outlets located above
the counter. preparation and storage rooms.
THE INTERSECTION OF HEALTHCARE
SIMULATION AND ICT Facility Telecommunications Infrastructure” standard
When designing structured cabling for a simulation space, on healthcare facility cabling.
consider that wireless access points and security cameras In addition to the cabling drops for the medical sys-
are commonly installed using power over Ethernet (PoE). tems, cabling is required for the simulation capture sys-
This impacts horizontal cable selection. Category 6A bal- tem. Ideally, all of the field simulation hardware should
anced twisted-pair cabling is typically recommended. The cable to the same room where the backend simulation
last worry someone should have when operating a simu- hardware lives, keeping the majority of data traffic local.
lation space or whose performance is being assessed While some connections to the healthcare simulation sys-
is whether the network will be unreliable. Therefore, tem are made with Category cables, like those for IP cam-
it is important that the cabling is suited to the demands eras, other connections, such as for speakers and micro-
that will be placed on it. phones, are commonly installed using AV cabling. Con-
Since simulation technology is rapidly advancing, sider sharing pathways. A single tray with a barrier or
it is worth considering not only futureproofing the a two-level tray can be advantageous. Sizing shared path-
cabling but also the pathways for the station cabling. ways is more complex due to different cable types and
Today what is served even by a Cat 6A cable may require sizes so proceed with care (Figure 9).
optical fiber in a future upgrade. If high
performing cable is used initially, cable
replacement may be delayed. If pathways
are designed with future cabling needs
in mind, those modifications become
much simpler.
Many simulation rooms emulate
an active medical space; therefore,
it is recommended to follow ANSI/
BICSI 004, Information Communication
Technology Systems Design and Implemen-
tation Best Practices for Healthcare Insti-
tutions and Facilities, Chapter 19 of
BICSI’s Telecommunications Distribution
Methods Manual (TDMM), 14th Edition
and other industry standards, such
as ANSI/TIA-1179-A, “Healthcare FIGURE 9: Two-tier basket tray awaiting cable installation.
If the AV design uses Category cable or connects
BICSI
to any network switches (including HDBaseT), then all
the cable and supporting hardware should be installed
STANDARDS
by the Division 27 cabling contractor. This may occur
by having the AV installer subcontract the structured
PRICELESS cabling contractor for installation. This helps retain con-

sistency among the structured cabling. It also allows the
CONTENT AT A same system warranty to apply to the structured cabling

dedicated to simulation as the other building cables.
ROLLED-BACK At a minimum, consider having the station cables for

the AV system meet the Division 27 specifications, even
PRICE if installed by a different contractor. This will allow the

owner to make moves, adds, and changes (MACs)
without requiring different materials.
BICSI’s International Standards Program
is a globally renowned, ANSI-accredited
standards development organization, and our MORE CABLING CONSIDERATIONS
best practices-based standards documents Additional cabling considerations include using color
discuss the design, installation, and integration coding to denote the different networks to which the
of information and communications technology cabling connects. This is extremely helpful to the simula-
(ICT) systems. Available standards include:
tion technical staff as long as it remains accurate. Unfor-
tunately, over time, color coding might not be followed,
• Educational Facilities
thereby negating its value. Labeling differences show the
• Building Information user which jacks are connected to the standard network
Modeling (BIM) and which are connected to the simulation network. The
• Healthcare Institutions ICT designer must work with the owner to implement
identification methods that will not be undone during
• Distributed Antenna
Systems (DAS) the first MAC because they are unclear or too complex.
Unique jack types may be beneficial in simulation
• Intelligent Buildings
areas, not because the traffic requires it, but because
• Wireless LAN in some scenarios, simulation devices are subject
to moving and a standard RJ45 plug may risk coming
BICSI Standards unplugged or function poorly due to the strain to which
is celebrating its
it could be subjected. A jack with a locking, ruggedized
Silver Anniversary
design may keep failure at an outlet or patch cord from
with special pricing
of these documents. occurring. It would clearly give a visual indication
of a non-typical network connection.
The flexibility of cabling should meet the flexibility
of the simulation space. Some simulation spaces are fitted
out with overhead grids to which power, network, and
other simulated utilities are attached. These may benefit
from the use of cable reels that are able to be relocated
Visit bicsi.org/standards as the simulated scenarios require different configura-
for pricing and to order today. tions. When cable reels are required, use care to select
a reel matching the cable type. Consider that the
72 I ICT TODAY
horizontal cable used with the reels can be for the same is usually only involved in the design of the simulation
system used elsewhere. However, often the reels them- spaces and not the supporting spaces, such as classroom,
selves need specialty cable capable of standing up to the meeting or office spaces, interaction with them may be
stresses of being reeled and unreeled. Since these reels limited. However, it is important to coordinate needs with
often require special ordering, it may benefit the owner the specialized equipment they plan to deploy.
to have a spare cable reel in case of failure. Cable ramps or overfloor raceway can be used to pre-
Floor boxes for simulation spaces should be gasketed vent trip hazards and protect cables extending from the
because the floors in these areas will require more fre- wall to the center of the room. Different ramp or raceway
quent mopping than a typical classroom. Floors in simu- sizes and configurations will be required depending
lation rooms can be subject to splashing of fake blood on the simulation because the pathway may have
and simulated medications. A unified floor box solution, to support more than just power and data.
where similar boxes are used in the simulation spaces ICT designers should be aware of the impact
as elsewhere in the building, simplifies future MACs. the pathways could have on the transfer of noise into
With the use of headwalls, booms, ceiling service and out of a simulation space, control room or debrief
columns and systems not seen in spaces outside of the room. Route pathways from a space where recording
medical realm, it is not uncommon to have unique is to occur to areas with less potential for noise. Avoid
cabling requirements within simulation spaces. ICT public areas, such as elevator lobbies or where the stu-
designers must work with the medical equipment consul- dents gather to prepare for simulations. Sealed conduits
tant to confirm outlet needs and elevations as depicted should be required instead of open cable baskets
in Figure 10. Since the medical equipment consultant to penetrate walls.
FIGURE 10: Simulated operating room as seen from control room/observation room.
WIRELESS FOR THE SIMULATION SYSTEM
Today, most high fidelity manikins connect wirelessly,
making wireless a critical piece of the simulation system. When locating the IT/AV
Using a hidden service set identifier (SSID) with authenti- equipment rooms serving the
cation is one way to protect the simulation data; QoS
to prioritize the traffic is common when the wireless simulation spaces, cabling density
access points also serve other SSIDs. For security reasons,
becomes an important
many institutions place a barrier between their wireless
network and their building network. However, with consideration.
healthcare simulation, there must be a way to connect
the wireless to wired ports in order to allow the manikin
data to be downloaded to the capture equipment.
Some manikins connect over Radio Frequency (RF) port on day one, changes can be made remotely without
but use a network connection to make the simulated the need to dispatch personnel.
monitor work. This style of manikin requires computers When Ms. Follmer was asked how ICT designers can
that support both a wireless and wired network connec- best meet the needs of the team running the simulation
tion at the same time. It is important to connect to both system, she indicated that early communication with sim-
a manikin’s wireless network and the wired simulation ulation center staff on network specific needs is import-
lab network, allowing for connection to additional simu- ant. Getting information about the devices and software
lation devices. The IT department may have to modify requirements will make the design process easier. Some
the standard group policy and network settings to meet devices will require a simulation center specific network
this requirement, since it is uncommon to allow two net- in addition to the traditional organization network. Some
work connections at the same time. When simulations devices may need static IP addresses set up to simplify the
are run repeatedly throughout the day, the total of the connection process. Firewall settings may need to be
simulations can outlast the charge of a battery. Therefore, modified so that information from the manikin system
some owners prefer to use hardwired connections for can pass to the monitors.
power and data. While not usually subject to Health Insurance Porta-
bility and Accountability Act (HIPAA) security rules, data
NETWORK CONSIDERATIONS—WHERE TO generated by a simulation system is subject to privacy
PLACE THE SWITCHES AND SIMULATION concerns. When using standardized patients, the systems
ELECTRONICS? may capture images and other data that should not be
There are important network considerations to be worked accessible by others. In any case, data integrity and net-
through when deploying a healthcare simulation system. work reliability are crucial because simulation environ-
Does the data ride over the same network and connect ments are high stakes and high stress. They immerse the
to the same switches as the rest of the building traffic? student in the medical encounter almost as if it were
Where should those switches and other simulation a real life or death situation. Students should be free
electronics reside? to focus on medical decisions and needed skills. They
In addition to coordination with the simulation should not be impacted by technical issues or concerned
center staff, the ICT designer must work closely with about the fate of the recordings after the simulation.
the institution’s IT department. A common network Complete isolation of the healthcare simulation system
can be beneficial when IT and other technologies merge. removes the instructor’s ability to access the captured
It is possible to provide isolation or connectivity through data outside of simulation areas, which severely limits
virtual LANs (VLANs), network access control lists (ACLs) their ability to assess the medical students. Other users
and firewalls. If every data outlet is connected to a switch throughout campus may even desire to watch live
74 I ICT TODAY
captures remotely. Therefore, a backbone connection switches should be automatically granted access to make
should connect the simulation backend to the campus changes to the simulation equipment.
network. This leads to a number of important questions. When locating the IT/AV equipment rooms serving
Where should switches serving the simulation system the simulation spaces, cabling density becomes an import-
be physically located? Should they be located in the tele- ant consideration. Simulation spaces, especially the sup-
com room (TR) or in a dedicated AV space? The case could porting rack room, tend to require large numbers of net-
be made to collocate or segregate. While there are cost work cables, and limiting these to a shorter distance, such
advantages to having the switches in a location that as a space directly adjacent to the simulation rooms, saves
is already limiting access and meets the required envi- on cable and pathway costs (Figure 11).
ronmental, grounding and power needs, some owners It also can lower the level of effort required to support
do not want AV personnel in their TRs. If the decision the spaces and reduce the risk of disruption when MAC
is made to collocate the simulation electronics in the work is occurring in other parts of the building. Depend-
TRs and security concerns exist, the use of locking cabi- ing on the building configuration, a location adjacent to
nets to control access to the various hardware should the simulation spaces may not allow the station cables for
be considered. the remainder of the floor to stay under the critical Ether-
No matter who has physical access to the healthcare net distance requirements (i.e., 100 m [328 ft] for Cate-
simulation equipment, it is important for the owner gory cables with the exception of Cat 8). One solution
to consider who will manage the switches supporting is to establish an AV room dedicated to the simulation
the system. Not everyone who has access to the building system but equipped as if it were another TR on the floor.
SIM CORRIDOR
C106
0B N2B N2B N7BN2B N10A

N12B
N2B
N23D
N20B SIM OR
113
SIM SERVER N4B
111 N12B N12B
N21A N2B N12B N2BN2B N2B N12B
FIGURE 11: Simulated operating room adjacent to server room.
To accomplish this, extend a fiber backbone from more traffic than would be desired around the
the MDF in the same fashion as to the other TRs. Connect simulation system equipment.
it to the grounding system and provide ladder racking Many providers of healthcare simulation systems
and other equipment. At this location also provide one offer cloud-based features related to storage and man-
or more remote edge switches, which allow convenient agement. These solutions support the on-site equipment
connections to the simulation equipment. This also rather than replace it. Remote management generally
allows horizontals requiring connection to the building appeals to the simulation support team; however, it is
LAN or simulation system to route to the same room, important to discuss cloud services with the owner’s IT
although there may be different cabinets or racks. department because it may run against institutional
Using the same switches as used for the building policy due to security concerns. While the discussion
network requires the owner to have less backstock in of collocation and cloud-based services is often led by
case of equipment failure. It also makes it more likely the team designing the simulation system, it is helpful
that the equipment will be monitored for faults and that to have the ICT designer participate in the discussion.
firmware and maintenance will be kept current. Consider They have insight into both the IT department stan-
a central uninterruptible power supply (UPS) for the dards and the project at large. Outcomes of this discussion
data/equipment room because it can also be used to pro- can impact ICT design, and it can be more challenging
tect simulation-related hardware if the area is prone
to power issues. Full body simulators can be expensive
and at risk of power related damage. A single power event
that causes the simulation or simulation computers
to reboot could break an entire student session.
Configuration of a collocated room is not as simple
as designing a larger TR. Simulation equipment requires
4-post racks or enclosed cabinets for equipment support,
which is not the case with most network electronics. The
need for different racks generally results in an irregular
row and eliminates the possibility of baying. Since cabi-
nets can be as deep as 42,” working clearances should
similarly scale up, thereby impacting room size. Coordi-
nate pathways in the room to support basic connections
as well as the healthcare simulation connections. Some
of the equipment in the room will be installed by the tele-
communications/ICT contractor while other hardware
is typically installed by the AV contractor. Plans should
show the space reserved for these trades. Coordination
of scope and timing by these trades is necessary to ensure
the room is laid out properly (Figure 12).
Although it may seem like a good idea to use the
control room to house the simulation equipment, this
is not recommended. The environment best suited for
electronics is not comfortable for most personnel. The
equipment generates a level of noise that would be prob-
lematic when a simulation is being run. Control rooms,
while having a certain level of security, generally have FIGURE 12: Simulation server room layout.
76 I ICT TODAY
to get clarification from the key players if any questions
are left unanswered.
An IT administrator with experience at an institution Today, most high fidelity
with a simulation center cautions ICT designers to care-
fully consider the size of the TRs, especially those that will
manikins connect wirelessly,
also house simulation equipment. With the evolution making wireless a critical piece
of the healthcare simulation systems, more equipment
will find its way into the TRs. Also, as additional simula-
of the simulation system.
tion technologies are adopted, network connections like-
wise increase. Flexibility to accommodate changes is of
vital importance. While this is true throughout other
spaces, it is critical in simulation centers. ICT designers plan changes, often their review occurs long after the
should also watch for value engineering. As cost manage- value engineering, thereby making it difficult to bring
ment exercises progress, the value of network infrastruc- the team back to the table to look for other cost
ture and cabling can assume less importance, especially saving opportunities.
when compared to the simulation systems. Since the IT In addition to making decisions about the simulation
staff is not always involved in the budget discussions, the system and the data network, there are considerations
ICT designer should seek to have them give feedback on how to approach simulation phones. There is the
on such reductions before implementing them. Although potential for liability if phones in the simulation area
the IT department usually has the opportunity to review only reach the control center and do not allow access
TestPro CV100
Multifunction Cable Tester
Test More. Test Faster. Test for Less.
Visit the
Need an Affordable, Upgradeable AEM Virtual
Alternative for Cable Certification? Booth
at the BICSI Fall
The TestPro delivers fast, accurate copper and fiber certification testing
Conference
for a fraction of what you’re paying now. Come by the AEM virtual
booth at the BICSI Fall Conference to learn more.
• Low-Cost Options Add Network, PoE, Multi-gig, and Wi-Fi Testing

• 2-Year Warranty
• Warranty Program Approved by All Leading Cable Manufacturers
2019
Silver
• Perfect for Cabling Contractors
Learn More at AEM-Test.com

NEW
PRODUCT
SHOWCASE
WINNER
Design, Diagnoistic, and
Installation Tools
AEM
For questions and to schedule a live demo,

2019 ISC WEST
contact us at CustomerCare@AEM-Test.com
AEM_BSCSI_Fall_VirtualAd_v2.indd 1 8/17/2020 11:05:06 PM
to 911. Some owners do not wish to use that many ports available as a 3D immersive experience for remote
in the voice over IP system or pay a provider for that users. Additional functionality, such as eye tracking of the
many phone lines. Some fear mobilizing emergency learner, is already available but is not yet widely adopted.
resources if a misdial occurs during a simulated code Virtual health, also known as telemedicine, means
blue. An option is to put signage at the simulation medical professionals no longer have to be in the same
phones and provide emergency phones at select room as their patients. Going beyond replacing office
locations in the corridor. visits, medical robots can be used to perform surgeries
and medical procedures and even allow a medical profes-
FUTURE OF HEALTHCARE SIMULATION sional to triage patients from outside of the medical facil-
Many factors are increasing the adoption of healthcare ity. This is an important benefit for underserved areas
simulation. Simulation, when paired with modified of the world or for those with rare medical conditions.
teaching methods, improves learners’ competencies. As technology increasingly makes its way into
Furthermore, they serve to reduce medical errors, which the delivery of medical care, it is not surprising that
benefit the medical profession and professionals, their the use of technology in medical education is increasing.
institutions, and most importantly their patients. There The future of quality medical care depends on reliable
is a need to increase patient safety at the same time networks designed, installed and supported by
as the aging population is taxing an already burdened ICT professionals.
healthcare system.
As technologies, such as virtual reality and aug- AUTHOR BIOGRAPHY: Patti Fisher, RCDD, DCDC, OSP,
mented reality, advance and become less cost-prohibitive, RTPM, WD is a technology and security designer for the Sextant
the use of healthcare simulation will continue to rise. Group, an NV5 company. A BICSI member for 20 years, Patti
These tools immerse learners in the educational experi- is a dedicated BICSI volunteer. She has facilitated the Greater
ence without the need to create an elaborate and expen- Kansas City BICSI Breakfast Club since 2005. She was active
sive scene. Not only is the implementation of simulation on the RCS Committee from 2007 until 2020. She currently
growing, the capability of simulation is also constantly holds the position of Secretary on the BICSI Standards Committee
evolving. Captured simulation information can be made and Vice Chair of the Codes subcommittee. Patti won the David K.
Blythe University of Kentucky Award for
Outstanding Member of the Year in 2012.
She is involved in Women in BICSI. She
thinks of herself as a BICSI ambassador,
Get real-time, challenging others to do what they can

to advance the association, industry and,
real-world learning in turn, their skill set. Patti can be reached
at pfisher@thesextantgroup.com.
with BICSI Virtual
Classroom Training.
bicsi.org/vct
78 I ICT TODAY
Maximize Your Productivity with
Installs.

Visit our virtual BICSI booth

September 27-October 1
americailsintech.com
AV Technology That Supports Hybrid Learning
Connect with Students in the Classroom or Online
Educators are facing unprecedented challenges right now. How can they teach their students at a
time when social distancing and masks are the rule in classrooms, and remote learning is the new
normal? Extron has technology to help instructors reach their students, in the classroom or online.
Voice Reinforcement for Instructors

The VoiceLift Pro microphone system allows instructors wearing a mask
to be heard by amplifying their natural speaking voice and distributing it
throughout the room and to remote learners.
Storage and Streaming for Remote Learning

SMP Streaming Media Processors record classroom sessions and
stream them online live or on-demand.
Remote Learner Interaction with the Classroom

The MediaPort 200 HDMI to USB bridge integrates the Classroom AV
system to soft codec and UC conferencing applications like Zoom, allowing
remote students to interact with those in the classroom as
if there in person.
Extron’s friendly and knowledgeable Education Support Team is always ready to assist with product or application
questions, technical support, and system design advice. Contact the Extron Education Support Team
at EducationSupport@extron.com or go to www.extronclassroom.com.

BICSI Ict-Today

Uploaded by

Copyright:

Available Formats

BICSI Ict-Today

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

BICSI Ict-Today

Uploaded by

Copyright:

Available Formats

ICT TODAY

THE OFFICIAL TRADE JOURNAL OF BICSI

A high-speed wireless highway for tomorrow’s

05 Essential ICT Professionals 36 Protecting the Future at the Edge: Addressing

46 Multimode Fiber Technology, Latest Market

66 The Impact of ICT on Healthcare Simulation:

26 Leveraging Infrastructure to Optimize the As health care simulation technology is increasingly

EDITORIAL REVIEW BOARD BICSI INFORMATION

Phone: +1 813.979.1991 Web: bicsi.org ICT TODAY NEEDS WRITERS

Jamie Finne, Creative Contact icttodayeditor@bicsi.org if you

Jeff Giarrizzo, Senior Technical Editor ADVERTISING SALES

A high-speed wireless highway for tomorrow’s

2/11/54 Mb/s 54 Mb/s 600 Mb/s 6.8 Gb/s 10 Gb/s 10 Gb/s

802.11 802.11 802.11 802.11 802.11 802.11

1997-1999 2003 2009 2014 2019 2020s

1980s 1990s 2000s 2010s 2020s

FIGURE 2: Wi-Fi 6E approvals. (Image: ©2020 CommScope. All rights reserved.)

ICT FIELD OF VISION SPONSORS

LEAD APPLICATIONS 6 GHz Backhaul

General Improvements in Existing Experiences

Licensed Versus Unlicensed Networks

the two technologies. THE Wi-Fi 6E OPPORTUNITY

UNDERSTANDING THE TERMINOLOGY

Li-Fi's OPERATIONAL LIGHT SPECTRUM caused by natural disruptions in daylight or sunlight.

Data Transfer Speed >1 Gb/s About 150 Mb/s

Security High Low

Power Consumption Potentially Less More

TABLE 1: Essential differences between Li-Fi and Wi-Fi.

Sales Representatives Are Waiting For Your Call!

FIGURE 3: An example of a Li-Fi installation.

Centralized Compute Architecture Distributed Compute Architecture

Edge deployments are no different than

Consistency: Edge is not a single

Remote Management: Intelligent infrastructure

Enclosures Cabling and Connectivity

 Magic Mirror is an independent free-

Visit bicsi.org/rcdd for details.

Data Processing Data Analytics Data Cashing

Edge Data Center

Radio Site Access Site Aggregation Core Site

Increasing size of an edge

FIGURE 2: Network-based edge locations size and latency.

PHYSICAL SECURITY CONSIDERATIONS Level 2: Electronic access with automated

Connect and Protect your Datacom Project with nVent CADDY

Learn more about nVent CADDY datacom solutions: go.nVent.com/datacom

©2020 nVent. H86282-USEN-2008

and making cooling more able ranges A1 to A4 as compared to the recommended

Communications for future, then developing and deploy-

to support the need for these newer

cooling methods and large-scale

Learn more at www.gofrontrow.com

in the New Normal?

800.422.6191 888.792.7463 800.847.5629 800.309.2322 801.484.5238 800.238.0787

100-300 m 500 m 2 km 10-20 km 40 km

1T deployed at a faster rate than

FIGURE 3: Information from FCIA’s Fibre Channel Roadmap.

Data Center Trends 19% CGAR 2016-2021

few years has been the accelerated migration to public

13% CGAR 2016-202

 Magic Mirror is an independent free-