5G/6G SPECIALFOCUS

FOCUS
SPECIALFOCUS TABLE OF CONTENTS

Cover Feature
4 
B eyond the Bench: Adapting Test and
Measurement for 6G in the Age of AI
Vincent Kotzsch, NI, Emerson Test & Measurement

Perspective
10 Artificial Intelligence In 6G: More Than
Large-Language Models
Roger Nichols, Keysight

Special Report
14 The Emergence of mmWave
Applications Drives Interconnection
Development
Pasternack, an Infinite Brand

Technical Feature
20 Paving the Way to Terabit Wireless
Telecommunications
Dr. John Howard, Rana Barsbai and Steve Jalil,
Electromagnetics Technology Industries, Inc.

Product Features
26 Python Package Controls AWG DDS
Multi-Tone Generation
Spectrum Instrumentation

32 MMIC Filter Design Goes Mainstream

with Web-Based Design Tools
Marki Microwave

Tech Brief
36 Bootstrap Oscillator Delivers
Exceptional Phase Noise Performance
Quantic Wenzel

Departments
38 New Products

STAFF
Group Director: Carl Sheffres Multimedia Staff Editor: Barbara Walsh EUROPE
Associate Publisher: Michael Hallman Electronic Marketing Manager: Chris Stanfa Office Manager: Nina Plesu
Media Director: Patrick Hindle Senior Digital Content Specialist: Lauren Tully CORPORATE STAFF
Brand & Content Director: Jennifer DiMarco Digital Content Specialist: Vincent Carrabino CEO: William M. Bazzy
Technical Editor: Eric Higham Director of Production & Distribution: President: Ivar Bazzy
Edward Kiessling Vice President: Jared Bazzy
Associate Technical Editor: Cliff Drubin
Editorial & Media Specialist: Kelley Roche Art Director: Janice Levenson
Associate Editor: Kaitlyn Joyner Graphic Designer: Ann Pierce

2 5G/6G  MAY 2024

Advanced Channel Emulation
for 5G NTN Networks

ACE9000 Advanced
Channel Emulator

The dBm ACE9000 can emulate

the effects of naturally occurring
disturbances in a non-terrestrial
communication environment.
Reducing the effects of these
impairments can extend network
reach, improve capacity and ensure
reliable connectivity even in geographically remote or under-populated hard
to reach areas, such as the open ocean or large, desert regions.

Link Emulation including:

Phase continuous delay, Doppler, and atmospheric scintillation

AWGN and Eb/No

12 tap (path) multipath fading with Rayleigh, Rician & CW.

Angle of Arrival (AOA), k-factor, and correlation controls

Hardware-in-the-Loop emulation

Learn More at
MAURYMW.COM
Vist us at IMS 2024
Booth #704. Your Calibration, Measurement & Modeling Solutions Partner!
 FOCUS
SPECIALFOCUS COVER FEATURE

Beyond the Bench: Adapting Test and

Measurement for 6G in the Age of AI
Vincent Kotzsch
NI, Emerson Test & Measurement, Austin, Texas

A
s the telecommuni- insights from measurement data to leading to an increased device un-
cations landscape help make business decisions. This der test (DUT) complexity requiring
gears up for the ad- strategic embrace of AI not only new testing methodologies.
vent of 6G technol- ensures seamless integration into
ogy, the role of artificial intelligence the dynamic wireless landscape but LEVERAGING GENERATIVE AI
(AI) becomes increasingly indis- also lays the foundation for future TO BOOST PRODUCTIVITY
pensable, bringing with it immense breakthroughs, propelling the in- As of 2024, wireless connectivity
opportunity for innovation as well as dustry toward a future defined by has become ubiquitous, reaching
major challenges for the wireless in- innovation and excellence. most households and mobile de-
dustry. On one hand, AI promises to vices worldwide. This widespread
usher in an era of improved mobile UNDERSTANDING THE adoption signals a shift in industry
network efficiency. However, testing INTERSECTION BETWEEN AI dynamics from a primary focus on
AI-integrated devices requires new AND 6G TEST revenue growth to one centered on
testing methodologies that push The convergence of AI and test- profitability and optimization. This
the boundaries of measurement sci- ing unfolds through a multifaceted transition towards optimization has
ence. On the other hand, as devices lens, encompassing diverse appli- propelled discussions surrounding
continue to increase in complex- cations and transformative poten- the integration of AI in 6G wireless
ity, meeting time-to-market, qual- tials. However, there are two main technology as industries seek to
ity and cost objectives will become use cases driving change in the in- enhance efficiency and maximize
increasingly complex. In this space, dustry. First, generative AI begins returns in this mature market land-
AI can help acquire more actionable to emerge as a tool for accelerating scape. In the context of test and
design workflows, measurement, AI-driven solutions
optimizing time- offer unparalleled opportunities for
Product Lifecycle

Product Complexity and Customer Demands More Tests to-market and streamlining workflows, optimizing
Competition and Shrinking Time to Market Less Time minimizing operat- resource allocation and enhanc-
ing costs. Second, ing overall operational efficiency.
Responding with Technology New Tools
wireless industry Figure 1 shows a product life cycle
Design Validation Manufacturing players are increas- flow, including some of the com-
ingly integrating AI pounding and conflicting issues
 Fig. 1 Product development life cycle. into their products, that pressure this product cycle.

4 5G/6G  MAY 2024

 FOCUS
SPECIALFOCUS COVER FEATURE

engineering efficiency within each

Fear their production processes Say they will lose market share step of the process. NI has devel-
57% are outdated and cannot keep
up with new business and
technology trends
46% within two years unless they make
significant changes to product
lifecycle processes.
oped a prototype of a significantly
optimized workflow driven by gen-
erative AI. This approach utilizes an
 Fig. 2 Findings from an NI 2022 research report. AI infrastructure with a chat interface
to autonomously generate tests and
Create
Develop Develop sequences based on given require-
Define
Outcomes
Test
and and Execute
Tests
Analyze
Results
ments and datasheets. Automating
Debug Debug
Plan
Tests Sequences this segment alone could slash the
<> time required for design character-
ization from weeks to days. The in-
Optimize Report tegration of tools like generative AI
Out
marks another significant leap for-
ward, promising to expedite time-
to-market, reduce operating costs,
$
enhance leverage and promote re-
usability.
Reduce Time
to Market
Deliver Customer
Satisfaction
Improve the
Bottom Line
Adapt to the
Future
AI-EMBEDDED DEVICES
INTRODUCE A NEW LEVEL OF
 Fig. 3 Simplified device characterization and validation workflow. COMPLEXITY
The integration of AI into 6G
The integration of generative A CASE FOR IMPLEMENTING wireless products ushers in a new
AI into design workflows marks AI IN SEMICONDUCTOR era of complexity and innovation.
a transformative leap for the in- WORKFLOWS Unlike their traditional counter-
dustry. By leveraging AI-driven in- Figure 3 shows a simplified ren- parts, AI-enhanced DUTs present a
sights derived from measurement dition of the workflow involved host of unique challenges that de-
data, businesses can unlock un- in characterizing and validating a mand innovative solutions to guar-
precedented levels of productivity device. The process typically com- antee both performance benefits
and competitiveness. Generative mences with defining the desired and trustworthiness. For instance,
AI algorithms play a crucial role in outcomes, which may stem from in telecommunications, AI-driven
optimizing the design of intricate various sources such as design network optimization algorithms are
antenna systems, enabling rapid specifications, product require- deployed to enhance spectral ef-
prototyping and iterative refine- ments or cost and time constraints. ficiency and minimize interference,
ment to meet stringent perfor- Engineers then translate these re- thereby elevating overall network
mance requirements. The need to quirements into a comprehensive performance and user experience.
realize these benefits is becoming test plan encompassing all neces- This widespread adoption of AI
much more concrete, as evidenced sary tests to evaluate the device underscores the pressing need for
by the data points shown in Figure against the specified criteria. Subse- the test and measurement indus-
2. These two results come from quently, they develop and refine the try to evolve and craft specialized
internal NI research conducted in tests, striving for maximum automa- solutions tailored to the testing re-
partnership with the research divi- tion. This phase often consumes the quirements of AI-enhanced DUTs,
sion of the Financial Times Group.1 most amount of test time. ensuring seamless integration and
This strategic convergence not Following test development, en- optimal performance in real-world
only expedites design workflows gineers execute the tests on mul- scenarios.
but also empowers engineers and tiple devices to assess individual As the industry gears up for the
designers to innovate. In sectors like performance and device-to-device era of 6G and beyond, complexity
the semiconductor industry, AI-driv- variations. The results are then me- reaches unprecedented levels as
en design tools revolutionize chip ticulously analyzed and reported devices integrate advanced func-
layout optimization, yielding higher with room for refinement or opti- tionalities and AI-driven intelli-
performance and energy efficiency mization of tests and sequences as gence. The dominance of software
while shortening design cycles. Ul- needed. Depending on the com- components over hardware further
timately, the integration of AI into plexity of the device, this process amplifies the demand for frequent
design processes presents unparal- may extend from weeks to months, testing to maintain reliability amidst
leled opportunities for enhancing exerting pressure to expedite it to rapid software evolution. The inte-
productivity, efficiency and innova- accelerate revenue generation. gration of AI exacerbates this com-
tion, propelling industries toward By leveraging state-of-the-art plexity, necessitating rigorous evalu-
sustained growth and competitive- software and hardware tools, like ation to guarantee safety and trust-
ness in the dynamic landscape of NI’s LabVIEW and TestStand, or- worthiness. Given the pivotal role of
wireless technology. ganizations can significantly boost these devices across industries, en-

5G/6G  MAY 2024 5

 FOCUS
SPECIALFOCUS COVER FEATURE

efficiently validate the performance

Data Test AI to Test AI and reliability of embedded AI tech-
• Sufficient Data Quality
and Quantity
• Scenario-Based, End-to-End
Test
• Reduce and Manage
Complexity of Test
nologies.
• Organization, Management, • Cyber-Physical Test Leveraging • Hyper-Automation of
Labeling, and Pre-Processing SIL, MIL, and HIL Approach Improvements to the MANAGING THE INFINITE TEST
Tools Based on Real-World Scenarios Model and Training Data SPACE
The “infinite test space” refers to
the vast array of potential scenarios
 Fig. 4 Embedded, trustworthy AI in 6G.
and conditions that an AI system
suring quality remains paramount, AI models, encompasses three main may encounter in real-world appli-
prompting new test challenges like steps: model design, training and cations. Unlike traditional software
efficiently sourcing (the right) data, validation. While synthetic training testing, where inputs and outputs
setting up accurate scenarios for data generated through simulation can be exhaustively enumerated,
test and managing the “infinite test tools offer some utility, the accu- AI systems are trained on data and
space.” Some details of these chal- racy of this training data hinges on may exhibit unexpected behaviors
lenges are shown in Figure 4. the fidelity of the simulation mod- when faced with novel situations.
While the technology of AI ap- els. In contrast, real-world training This means that testing AI involves
plied to wireless is new and with datasets, acquired under authen- grappling with an expansive and of-
many challenges, addressing them tic channel conditions, offer supe- ten unpredictable range of circum-
allows the wireless industry to ef- rior quality but are inherently more stances, making it impractical to
fectively harness the transformative challenging to obtain due to the test every possible scenario. Hier-
power of embedded AI. In applica- specialized hardware and software archical scenario descriptions, cou-
tions where spectrum, energy and required for data recording. Striking pled with smart scenario reduction
chip real estate are both finite and a balance between these consider- techniques, are needed to manage
valuable resources, AI can help op- ations is essential to ensuring the ro- the complexity of the test and mea-
timize them in next-generation wire- bustness and reliability of AI models surement process while maintaining
less devices. This helps pave the tailored for 6G applications. the test coverage and ensuring the
way for enhanced performance, re- reliability and robustness of an AI-
liability and efficiency in future wire- SCENARIO-BASED TESTING: enhanced DUT.
less networks. THE FUTURE FOR AI-
EMBEDDED DUTS IMPLICATIONS FOR FUTURE 5G
SOURCING DATA EFFICIENTLY: Traditional stimulus-response & BEYOND DEPLOYMENTS
BALANCING QUANTITY AND testing systems fall short when it The integration of AI into wireless
QUALITY comes to validating embedded AI applications represents a pivotal ad-
The integration of AI into test devices. The challenge lies in the vancement, with discussions around
and measurement processes stands unpredictable nature of machine its role in future standards gaining
at the forefront of a transformative learning-trained systems, which momentum. Notably, the 3rd Gen-
era, poised to redefine industry may exhibit unexpected behavior eration Partnership Project (3GPP)
standards and methodologies. This across a broad spectrum of test sce- is actively exploring AI’s integration
is particularly true in the realm of 6G narios. Unlike traditional algorithms, into the forthcoming 5G Advanced
wireless and beyond. This strate- embedded AI models can be highly and 6G standards, a topic of signifi-
gic embrace of AI not only ensures sensitive to environmental factors cant interest in Release 18 and 19
seamless integration into the dy- and system configurations, making discussions. This strategic embrace
namic wireless landscape but also it challenging to identify relevant of AI is not merely speculative but
lays the foundation for future break- stimulus signals. To address this, grounded in pragmatic consider-
throughs, propelling the industry to- scenario-based test systems of- ations, driven by the imperative
ward a future defined by innovation fer a viable alternative, abstracting to enhance profitability within the
and excellence. different test conditions into easily wireless industry. Given the scarcity
AI algorithms are poised to play understandable scenarios with pre- of resources such as spectrum, size
a pivotal role in revolutionizing test defined parameters and outcomes. constraints and power consumption
management processes, automat- These scenarios are then dissected concerns, even marginal improve-
ing scenario selection, optimizing into concrete test cases for compre- ments facilitated by AI can yield
test coverage and mitigating the hensive evaluation. Despite the po- substantial cost savings. AI holds
complexities associated with di- tential complexity of scenario selec- the potential to unlock significant
verse use cases and edge condi- tion, leveraging smart techniques gains across various fronts, includ-
tions. However, testing AI devices enables the identification of perti- ing spectral efficiency, interference
in the context of 6G applications nent scenarios, ensuring thorough reduction, chip size optimization
presents formidable challenges, testing of AI devices in the wireless and power consumption, thereby
particularly in the critical phase of AI space. As the wireless landscape reshaping the landscape of wireless
model training. This phase, essen- continues to evolve, scenario-based technology. Figure 5 shows a base
tial for creating robust and reliable testing is a promising approach to station tower and sectors, along

6 5G/6G  MAY 2024

 F nd R P

TM
MAGPro
R

Coilcraft’s MAGPro suite of online inductor a full range of performance graphs including
analysis tools are designed to enable L, Q and ESR at frequency.
inductor selection and circuit optimization For RF choke applications, the Z @ Frequency
based on sound engineering principles and tool searches thousands of part numbers for
measured data. your desired impedance and lets you select
The RF Inductor/Choke Finder and Analyzer the desired choke based on size, performance,
offers two search options. The L @ Frequency or a combination of the two.
option identifies inductors suitable for your Reduce your design cycle time with
L, Q, and current requirements and displays confidence at www.coilcraft.com/tools.

WWW.COILCRAFT.COM
 FOCUS
SPECIALFOCUS COVER FEATURE

be met. These include access to a As we move closer to the era

sufficient quality and quantity of of 6G, collaboration, standardiza-
training data, the development of tion and the development of AI-
test systems capable of emulating powered testing solutions will be
real-world scenarios and the imple- critical. Establishing industry-wide
mentation of robust methodologies standards for testing 6G devices
for navigating the infinite array of with embedded AI is paramount.
testing scenarios. These standards should encom-
pass AI algorithms, AI training and
THE BOTTOM LINE: AI WILL corresponding data, as well as per-
 Fig. 5 Wireless base station tower, REVOLUTIONIZE THE WIRELESS formance metrics and testing meth-
sectors and backhaul installation.
SECTOR odologies, to ensure consistency
with wireless backhaul links, as an The integration of AI into the and reliability. Collaboration among
example of a network that can ben- physical layer of 6G devices repre- device manufacturers, network op-
efit from AI integration. sents a monumental leap in wireless erators and AI experts is essential
While AI’s current application pri- communication technology, promis- to address the challenges of AI inte-
marily targets higher network layers, ing to unlock new possibilities and gration. Sharing best practices and
its potential extends to lower layers transform industries. However, with insights can lead to more robust
of the protocol stack, presenting a great innovation come significant testing methodologies, as well as
burgeoning trend in 6G discourse. new challenges, particularly in the ensuring interoperability with opti-
Research endeavors are actively realm of testing. The industry must mum performance. By addressing
exploring the integration of AI into proactively address the test implica- these challenges head-on, we can
lower layers, recognizing its capac- tions of AI integration to ensure that help ensure that AI-integrated 6G
ity to improve spectrum and energy 6G devices deliver on their prom- devices operate flawlessly, ushering
efficiency as well as performance. ises of ultra-low latency, reliabil- in a new era of connectivity.
However, to realize the promise of ity, energy and spectrum efficiency,
embedded AI within wireless net- massive connectivity and blistering Reference
data rates. 1. INSIGHTS.NI.COM/DESIGNED-TO-
works, several prerequisites must PERFORM/

R&S®NRP170TWG(N)
Thermal power sensor

Discover the R&S®NRP170TWG(N) thermal power sensor, the market‘s only

NMI-traceable power sensor for precise D band measurements, covering
110 to 170 GHz. Engineered for simplicity and top performance, it integrates
into any setup. Moreover, with fast, accurate and stable readings, as well
as USB/LAN connectivity, it provides capabilities essential for 6G R&D
and advanced radar applications, allowing you to measure power with
confidence - and exceed expectations.
Accelerate engineering
efficiency through test.
ni.com/semiconductor

Visit us at Booth
1839 at IMS
 FOCUS
SPECIALFOCUS PERSPECTIVE

Artificial
Intelligence In 6G:
More Than Large-
Language Models
Roger Nichols
Keysight

T
he late 2022 launch of OpenAI’s ChatGPT
and subsequent enhancements through
GPT-4, changed the world’s perception of
the maturity and potential of artificial in-
telligence (AI). Governments are now writing regula-
tions, industry is developing technology and business
models and academics are probing the latest research
topics. Driven by a chatbot built on a large-language
model (LLM) created by the transformer architecture,
this tidal wave of activity is hiding many practical AI
developments that are more relevant to radio systems
in mobile wireless.
As early as 2019, the ITU-T’s Network 2030 Focus
Group (FG Network-2030)1 highlighted the necessity
of AI from the physical to the application layer of the
6th generation of wireless networks to accommodate
the demands of realizing their vision. Thus, the 6G vi-
sion has always included AI as a fundamental building
block and tool. However, LLM’s AI models, trained on
the vast database of text on the global internet, are not
the solution for overcoming many technical challenges
in wireless communications. Since FG Network-2030,
there have been myriad journal articles, research pa-
pers, technical demonstrations, early standards work
and commercial solutions illustrating that the intensity
of this work is focused on machine learning (ML) unre-
lated to LLMs. In most cases, language models will not
be adequate for wireless and particularly radio tech-
nology. Instead, they will require models trained on
other sources of data such as radio I/Q pairs, signaling
traffic or user (payload) data.
ML-based optimization is the subject of research
and development at all layers of the wireless network.
Allow me to examine just a snapshot of the work clos-
est to the physical layer. This includes AI as applied to
the “air interface.” An excellent early overview2 delin-
eated novel concepts using ML to:

10 5G/6G  MAY 2024

 FOCUS
SPECIALFOCUS PERSPECTIVE

• Create custom waveform and Data and model validation: designing. Rather than becoming
modulation optimized for trans- Training AI models requires tremen- cynical or worried, I view these as
ceiver hardware and real-time dous amounts of data that fits the technical challenges characteristic
channel conditions balance of being “random enough” of any advance in technology that
• Design and train transceivers (unbiased, uncorrelated) while also our engineering community has
themselves based on use-case being “appropriate enough” (rel- put to good use time and again. I
specifics evant to solving the problem at have seen the interesting results of
• Allow these capabilities to moot hand). Clean and controlled training ML-designed filters and antennas
PHY- and MAC-layer standards data has thus become a premium and, while not always practical, they
themselves by standardizing only commodity. While LLMs can capital- change one’s perspective as to how
how ML would be implemented ize on the vastness of internet data, to meet the demanding key perfor-
for such bespoke and real-time data for solving more specific tech- mance indicators of our industry. In
determination of the air interface nical systems issues is less plentiful a demanding technical environment
• Take steps towards semantic and often private and proprietary. like radio systems and wireless com-
communications rather than opti- These two words are operative in munications networks, there is much
mizing only how to transmit bits. the implications of how the data work ahead of us to not only vali-
A panel of experts in 20213 were can be used. And, once a dataset date models and datasets but also
asked if AI was already used in con- is available, how does one know to validate and improve the results
temporary commercial wireless and whether it is adequate, appropriate, of the AI-optimized behavior and
the answer was a resounding “yes.” unbiased and secure? Once a mod- designs themselves. The combina-
AI was already being used in 4G el is trained, designers learn that the tion of conventional and AI-enabled
and 5G applications including traf- model itself requires continuous im- means of such measurement is an
fic load balancing and signaling provements and thus, model valida- intriguing and exciting area of de-
optimization, MIMO precoding al- tion has become a critical step. velopment and I am looking forward
gorithms, energy-use management Standards: Perhaps the most rel- to working to make the most of this
and network planning. The intent is evant work for wireless is in 3GPP.5 technology.
to expand the use of ML to drive im- 3GPP started AI standardization dis-
provements in system behavior that cussions as early as Release-17 with References
have become so complex that con- RAN3 initial study items6 related 1. Focus Group on Technologies for Net-
work 2030, Web: www.itu.int/pub/t-fg-
ventional solutions are constrained to data collection and focusing on net2030.
by their deterministic one-size-fits- energy saving, load balancing and 2. J. Hoydis, F. A. Aoudia, A. Valcarce and
all mathematical models. mobility optimization. RAN1, in Re- H. Viswanathan, “Toward a 6G AI-native
For ML to drive large-scale reli- lease-18, added an extensive study Air Interface,” IEEE Commun. Mag., Vol.
59, No. 5, pp. 76–81, May 2021.
able and viable improvements in item on using ML for improving 3. “Exploring the Role of AI in Wireless,”
performance, quality of service channel-state information, beam Keysight, 2022, Web: www.youtube.
and even quality of experience, the management and positioning accu- com/watch?v=DW0iZyXcDr8.
challenges we must address fall into racy. Work has progressed to mul- 4. “The Turbulent Past and Uncertain Fu-
three categories: tiple normative work items in 3GPP ture of Artificial Intelligence,” IEEE Spec-
trum, Sept. 2021, Web: spectrum.ieee.
Known weaknesses: The Octo- as part of Release-19, now under- org/magazine/2021/october.
ber 2021 issue of IEEE Spectrum4 way. 5. Finding AI in 3GPP, 3GPP, August 2022,
featured a cover asking: “Why is AI This all must happen in the con- Web: www.3gpp.org/technologies/find-
so Dumb?” Charles Q. Choi’s article text of governments developing as- ing-ai-in-3gpp.
6. “Study on Enhancement for Data Col-
therein described seven “ways AI’s sociated policies related to AI tech- lection for NR and ENDC,” 3GPP, Web:
fail.” These included brittleness, nology. Related headlines include portal.3gpp.org/desktopmodules/
embedded bias, catastrophic for- European Parliament’s landmark Specifications/SpecificationDetails.
getting and perhaps the most chal- law7 related to proper use, security aspx?specificationId=3817.
lenging issue: a lack of both “ex- and consumer recourse, the U.S. Ex- 7. “Artificial Intelligence Act: MEPs
Adopt Landmark Law,” European
plainability” and “common sense.” ecutive Order on AI Safety and Se- Parliament, March 2024, Web: www.
I have read about recent progress curity8 and the subsequent founding europarl.europa.eu/news/en/press-
in addressing the former but neu- of the U.S. AI Safety Institute.9 Much room/20240308IPR19015/artificial-intel-
ral-network ML suffers from a lack of this policy work is focused on the ligence-act-meps-adopt-landmark-law.
8. “Executive Order on the Safe, Se-
of explainability when answers are impact of LLMs on the internet and cure, and Trustworthy Development
“right” or “wrong.” From an engi- other media as well as on the secu- and Use of Artificial Intelligence,” The
neering perspective, understanding rity of critical communications and White House, Oct. 2023, Web: www.
the “whys” is essential to reliable compute infrastructure. We can ex- whitehouse.gov/briefing-room/presi-
and viable solutions. One can see pect impacts on the deeper techni- dential-actions/2023/10/30/executive-
order-on-the-safe-secure-and-trustwor-
“common sense” manifest daily in cal uses of AI in unpredictable ways. thy-development-and-use-of-artificial-
the inanity of AI-generated news For radio engineers, we are al- intelligence/.
articles and some of the false and ready seeing new approaches to us- 9. U.S. AI Safety Institute, Web: www.nist.
ridiculous answers to questions ing AI to not only manage wireless gov/artificial-intelligence/artificial-intelli-
gence-safety-institute.
posed to LLMs. communications but even do some

12 5G/6G  MAY 2024

 OTA ANTENNA TEST SYSTEM
This over-the-air (OTA) antenna test system is offered in partnership with MilliBox and Eravant. Each OTA bundle is
anchored by CMT’s 9 GHz VNA and includes a benchtop compact anechoic chamber, frequency extender modules,
a 3D antenna positioner (Gimbal), sophisticated but intuitive measurement software developed by MilliBox, and other
necessary accessories to fully set up a chamber for far-field and sub-THz antenna measurements.
Learn more about the new OTA Antenna Test System with the QR code below.

Sign up for a free virtual demonstration today and experience value firsthand.

®
Visit CMT at IMS 2024 and See This
Solution in Person - Booth #1111
 FOCUS
SPECIALFOCUS SPECIAL REPORT

The Emergence of mmWave

Applications Drives Interconnection
Development
Pasternack, an Infinite Brand
Irvine, Calif.

I
n recent years, a surge of applications, like satellite television industry exited the 2000s, mass
new applications, along with and marine or aerospace commu- manufacturing of mmWave chipsets
legacy applications, are uti- nication platforms, used mmWave, and integrated circuits (ICs) started
lizing the mmWave frequen- but terrestrial wireless communica- becoming more commonplace. De-
cies. This has been enabled by the tions applications were still in their velopments enabled cheaper and
availability of lower-cost mmWave infancy. During this era, 2G cellular more accessible 24 GHz automotive
semiconductors and the advent of speeds were a fraction of a megabit radars that were far less expensive
active/advanced antenna system per second (Mbps). than the previous mmWave tech-
(AAS) technologies for communica- After the turn of the century, con- nology that, largely, served defense
tions and sensing that are viable for sumer, commercial and industrial and government satcom applica-
consumer, commercial and indus- wireless communication solutions tions.
trial applications. The expansion of started gaining traction and auto- In 2009, the WiGig Alliance intro-
mmWave technology and applica- motive radar operating in the 24 duced WiGig, which was designed
tions is generating new challenges GHz band became available. Early to operate in the 60 GHz band.
with production tests. Traditional mmWave automotive radars still This technology was intended as a
mmWave test methods have been operate in the 24.0 to 24.25 GHz in- wireless standard to replace cable
based on labor-intensive manual dustrial, scientific and medical (ISM) in home theater and wireless dock-
techniques that require relatively band, which is sometimes referred ing for mobile device applications.
long unit test times. These new to as narrowband (NB). The NB au- These applications started becom-
applications are creating substan- tomotive radar application is limited ing popular with the advent of the
tial pressure for modern mmWave in use due to the narrow bandwidth smartphone and 3G. WiGig was
production and quality tests to be but is still viable for automotive subsequently absorbed into the
performed faster and at lower costs. emergency braking and adaptive Wi-Fi Alliance and while there were
This is causing a shift in the style of cruise control. Despite the NB na- limited launches of WiGig hardware
interconnect used for testing and ture of 24 GHz automotive appli- and 60 GHz Wi-Fi routers, the stan-
the need for greater levels of auto- cations, the mmWave spectrum is dard did not experience commer-
mation. characterized by wider bandwidth cial success. Wi-Fi efforts are now
channels and backhaul applications directed toward lower frequency
INCREASING AVAILABILITY OF make use of this feature in unli- IEEE 802.11ax (Wi-Fi 6) and emerg-
MMWAVE TECHNOLOGY censed and licensed bands. ing Wi-Fi 7 applications. One of the
Before the 2000s, the opportuni- Despite several applications, culprits for this lack of commercial
ties for mmWave technology were much of this early hardware was success was likely the higher cost of
primarily in defense, government manufactured in small batches us- 60 GHz WiGig chipsets. The high
or aerospace applications. Some ing manual fabrication, assembly cost of the chipsets discouraged
satellite communications (satcom) and testing techniques. As the designers from integrating 60 GHz

14 5G/6G  MAY 2024

 FOCUS
SPECIALFOCUS SPECIAL
COVER FEATURE
REPORT

Wi-Fi features into user devices. Fig- architectures that rely on one high- other technologies. Figure 2 shows
ure 1 shows an E-Band radio block power mmWave signal path. Lower- a 77 GHz radar block diagram.
diagram and the need for high per- power ICs provide the desired result
formance, high frequency, cost-ef- in conjunction with beamforming ar- MMWAVE TRENDS INFLUENCE
fective RF components is clear. chitectures that utilize antenna driv- RF INTERCONNECT
In the early 2010s, Ku-Band ICs ers. MIMO methods complement As described, the emergence of
became more readily available beamforming architecture to enable mmWave technology in non-gov-
and the race toward low earth or- smaller, more compact mmWave ernment applications is happen-
bit (LEO) satellite constellations antenna systems. This approach ing quickly. The rapid adoption of
for global broadband began. As also benefits from the smaller physi- mmWave technologies into these
Ka-Band satcom networks became cal sizes of mmWave antenna ele- applications relies heavily on the
more prevalent, investment and in- ments. development of mmWave chipsets
terest in Ka-Band satellite constel- After 2015, the 3GPP standard and advances in computing and
lations and ground terminals for adopted mmWave spectrum use simulation software. However, the
commercial applications increased. and Ka-Band satcom chipsets be- mmWave interconnect ecosystem,
During this period, mmWave fre- came more widely available. In ad- which has been in place for decades
quencies emerged as a key enabler dition to wireless mobile communi- in response to existing government
for the 5G vision. cations, fixed-wireless access (FWA) and satcom requirements, is evolv-
The development of phased ar- and other satellite broadband appli- ing to meet the needs of these new
ray antennas and antenna array-re- cations that make use of mmWave mmWave applications.
lated technologies has been a big frequencies have emerged. Now, To enable these emerging
factor in making mmWave wireless a host of mmWave applications, mmWave applications, interconnect
applications attractive for consumer, such as 5G, satcom, Wi-Fi and 77 density is increasing, along with
commercial and industrial applica- GHz automotive radar, incorporate operating frequencies. Supporting
tions. These technologies and ar- mmWave frequencies and compo- these trends is necessary to meet
chitectures require more RF paths, nents. There is a broad portfolio of the performance requirements of
but these paths operate at lower applications and use cases that are mmWave antenna array applica-
RF powers. This approach provides benefiting from the heavy invest- tions and AAS systems that have
advantages compared to traditional ment into mmWave chipsets and many more signal paths than legacy
mmWave systems. As described,
IF1 the signal paths in these new archi-
Mixer
tectures are at lower power levels
Mixers
Attenuator Passive than single-path systems. This en-
PA PA PA BPF LNA
Attenuator
Hybrid IF2 ables semiconductor technologies,
Tx
Antenna 71/81 GHz
PA like silicon, GaAs, InP and GaN to
PA provide the required performance.
x3 An important performance con-
PLL
Coupler sideration in these applications is
Rx
71/81 GHz
x2 VCO
thermal dissipation. Conduction
LNA Mixer LNA PLL
and dielectric losses are a function
Coupler of frequency, meaning intrinsic loss-
VCO
Attenuator es in a signal path are much higher
for mmWave frequencies. Using
PA multiple signal paths decreases
x3 x2
the power and dissipation in each
PLL
Coupler path. This makes it possible to use
LNA VCO
a distributed thermal management
approach in an mmWave system
 Fig. 1 71/81 GHz E-Band radio block diagram. Source: Pasternack.
design that accommodates the in-
creased number of signal paths.
19 GHz 19 GHz/38 GHz Doubler 38 GHz/77 GHz Doubler/MPA
VCO This approach avoids a high con-
centration of thermal energy in a
PLL x2 x2 Tx
Coupler small region and minimizes the ther-
77 GHz SP3T mal design challenges.
÷N In addition to an increase in the
Prescaler
interconnect volume for devices
MPA
Base
and systems, mmWave intercon-
Band
Rx nects require better tolerances to
IQ Mixer LNA ensure proper connection. The op-
77 GHz SP4T
eration of a transmission line and
 Fig. 2 77 GHz automotive radar front-end block diagram. Source: Pasternack. the interaction of electromagnetic

5G/6G 
5G/6G 2024
MAY 2024
 MAY 15
15
 FOCUS
SPECIALFOCUS SPECIAL
COVER FEATURE
REPORT

brations. Rigid and

PCB 1 semi-rigid coaxial
cable assemblies
generally cannot
PE44944 be reworked and
SMP Smooth Bore
should be formed
and left in place.
PE9776
SMP F-F Long Using these cables
Bullet Adapter requires some
planning, along
PE44946 with adequate
SMP Limited tools and exper-
Detent
tise in the labora-
PCB 2 tory or test floor to
properly form and
 Fig. 3 Manual test probe positioner.
install the con-
 Fig. 4 Blind mate coaxial interconnect.
nectors. Figure 4
waves within a transmission medi-
um is a function of frequency. This manual processes for prototyping shows a blind mate coaxial connec-
means that the required surface and quality testing do not support tor that uses a bullet adapter to con-
finish, feature tolerances and size the time and cost requirements. nect between two PCBs.
of the transmission lines scale with The result is a growing use of spring Another solution is to use probes
frequency. At higher frequencies, probes (pogo pins), coaxial spring and precision positioners to make RF
better finishes, machining and tol- probes and blind mate connectors, connections for prototype and veri-
erances on coaxial alignment result as opposed to the legacy standard fication/quality testing. Using probe
in the lowest loss and VSWR, along threaded coaxial connectors and interconnects in this way requires
with the best conformability for the solderable coaxial cable intercon- planning and design to ensure that
coaxial cable assembly. nects. Figure 3 shows a manual there are enough test points for the
Higher frequency interconnects test probe positioner and ground- RF paths. This test setup requires
must also have better overall align- signal-ground probe with a 2.92 complex probe heads and precision
ment tolerances than lower frequen- mm coaxial connector interface that positioning to connect every test
cy cables to ensure optimal perfor- may be used for mmWave test ap- point properly and reliably. These
mance. For instance, the wavelength plications. tests may require several to thou-
of a 500 MHz signal is roughly 600 RF port counts are increasing rap- sands of test cycles, depending on
mm, a 50 GHz signal is 6 mm and a idly. Systems that have traditionally the device or system under test. In
150 GHz signal is 2 mm. As a rule of incorporated roughly one to four some cases, it may be necessary to
thumb, it is recommended to keep ports are now routinely incorporat- probe multiple sides of a device or
feature resolution and tolerances ing 64 or more ports. This increases system simultaneously. As an exam-
below one-tenth of a wavelength the urgency of finding smaller, low- ple, many transmit-receive modules
and ideally below one-twentieth of er-pitch, high performance inter- for AAS applications are fabricated
the highest operating wavelength. connects for mmWave applications. on planar laminates or ceramic
For the 500 MHz feature resolution In many cases, flexible coaxial ca- boards where the routing may only
just described, tolerances should be ble assemblies that are commonly allow probe test points to be on
below 60 mm and ideally, below 30 used in test and measurement ap- the top and bottom. This requires
mm. At 50 GHz, the resolution/toler- plications to accommodate various dual-sided probing to access all test
ances should be below 600 μm and devices under test and test setups points and typical probe stations/
ideally below 300 μm. are not adequate for mmWave ap- probe positioners are designed to
Within an RF module and an plications. Rigid and semi-rigid co- probe only a single side. This may
AAS, it is possible to use planar axial cable assemblies tend to have require a custom probe station or
transmission lines, vias, high fre- lower losses and better VSWR than precision probe positioner. Figure 5
quency mezzanine connectors, sol- comparable flexible coaxial cables. shows a blind mate connector op-
derable coaxial cables and other However, skilled technicians must erating from DC to 22 GHz that can
high-density, board-to-board and shape rigid and semi-rigid coaxial be used in these applications.
cable-to-board interconnects that cables to properly install these in-
present a relatively small pitch and terconnects. Flexible coaxial as- AUTOMATION IS
profile. However, for test and mea- semblies only require a technician INCREASINGLY CRUCIAL
surement applications, like proto- to push-fit or properly thread and There is a shift away from high-mix,
typing and quality control, the types torque a threaded coaxial connec- low volume mmWave devices and
of interconnects just described are tor. These assemblies do, however, systems to much higher volume pro-
not generally viable. With mmWave require some attention to ensure duction. Part of this volume increase
technology being used in more and minimal deflection of the flex cable has been enabled by greater levels of
higher volume applications, long, during operation or between cali- integration of mmWave technology.

16 5G/6G
5G/6G
MAY
MAY2024
2024
Telecom and 5G/6G applications
Telecommunications is a rapidly changing industry; the constant march toward higher
frequencies, greater bandwidths, and faster data rates quickly makes yesterday's engineering
solutions obsolete. As the RF spectrum gets more crowded, one must manage unwanted EM
noise with the careful application of RF absorbing materials.

At PPG, engineers at our Cuming Microwave business have a breadth of experience with the
issues of noise suppression, resonance elimination, passive intermodulation problems, and
antenna sidelobe suppression. We can recommend the correct absorber material for your
application and help you quickly solve your RF noise problems.

Products:
• Millimeter wave absorbers
• Thin, cavity resonance absorbers
• Low-loss dielectric materials
• Anechoic chambers and other test environments
• Conductive materials
• Custom formulations and configurations available

Visit us at IMS Microwave Week 2024, Washington, DC

18 - 20 June 2024, Booth 1746

264 Bodwell Street • Avon, MA, USA • Tel: 508-521-6700

cmcsales@ppg.com • cumingmicrowave.com

Cuming Microwave is an ISO 9001:2015 manufacturer.

 FOCUS
SPECIALFOCUS SPECIAL
COVER FEATURE
REPORT

very high, the repeatability of blind

mate connectors and other inser-
tion methods is less so. The speed of
mating and occupied space advan-
tages of blind mate connections may
outweigh repeatability concerns.
The optimum solution is to enhance
the repeatability of blind mate con-
 Fig. 6 Calibrated noise source nections with robotic connection cy-
module. cling instead of human operators. A
ous test setups. Beyond small-signal robotic insertion system will likely en-
S-parameter testing, there is also a sure much better repeatability than
 Fig. 5 Pasternack BMA connector.
need for large-signal S-parameters human operators for spring probes
Chipsets and ICs with more integrat- and power, impedance (load-pull) and blind mate connections. The
ed features are available but there and/or noise testing measurements goal would be to achieve the same
also has been a shift toward more in many mmWave systems, espe- repeatability as threaded coaxial
automation and less manual manu- cially TR modules and other active connectors. Using spring probes and
facturing, assembly and testing. Au- RF devices and systems. Tradi- blind mate connectors with automa-
tomating manufacturing and assem- tional quality/verification testing of tion can also reduce the overall test
bly of RF parts has been an ongoing mmWave devices and systems usu- time and may reduce the test system
advancement for decades and auto- ally involves the use of various test footprint as robotic systems can be
mation in this area leverages many of stations or lines. Naturally, having designed that require less area than
the technologies used for high speed to disconnect, transport and recon- human operators need. Faster inter-
digital and computing systems. How- nect a device or system under test connect speeds could result in high-
ever, automation of both prototype in a different location, even within er throughput that would lower test
and laboratory RF testing and quality/ the same facility, can result in test costs and increase the ROI of what
verification testing of mmWave com- variations and changes. Consisten- are becoming increasingly expensive
ponents and systems has lagged oth- cy and repeatability concerns result- test and measurement systems for
er automation implementations that ing from the various test domains emerging high port count mmWave
have realized lower costs and higher of a given device or system under systems.
volumes. test make harmonizing the results
Initial applications have not lent of these test domains difficult, if CONCLUSION
themselves to automating mmWave not impossible. Figure 6 shows a For decades, mmWave tech-
tests. Government and defense precision-calibrated noise source nology was relegated to defense,
agencies have dominated these ap- module with 2.92 mm coaxial con- aerospace, space and some back-
plications and they can accommo- nector interfaces that may be part of haul communications applications.
date the costs and yields of manual an mmWave test setup. To support the insatiable demand
and time-consuming mmWave test Making the test setup flexible for data consumption, along with
strategies. Until the recent upticks enough to perform all the necessary mitigating spectrum clutter con-
in volume and cost concerns, there quality/verification tests minimizes cerns in the sub-6 GHz telecom
has not been substantial pressure to data repeatability concerns. It also spectrum, applications are mov-
develop lower-cost, faster and more results in higher test throughput ing to higher frequencies. The ad-
automated solutions. Additionally, and fewer steps with less manual vent of mmWave 5G applications
government and defense users may intervention. As an example, this has renewed interest in consumer,
be more willing to sacrifice repeat- approach might allow small-signal commercial and industrial mmWave
ability and cost in the products to and large-signal S-parameters for a communications and sensing. This
meet stringent standards and per- transmitter module to be measured interest is being kindled into a
formance requirements. and the data harmonized with- roaring flame with many new non-
Technologies like electronic cali- out substantial manipulation. This government mmWave applications
bration (e-Cal) have been instru- would provide a more complete emerging and evolving. Some of
mental in making typical small-sig- analysis of the overall transmitter these applications include 5G FR2,
nal RF testing more repeatable and behavior for sensing or communica- 802.11ay from 57 to 64 GHz, au-
reliable. However, these technolo- tions applications. tomotive radar from 77 to 81 GHz,
gies still require interconnect cy- Another outcome of additional mmWave imaging/radar for secu-
cling between calibrations. Greater automation, especially with intercon- rity and machine learning and new
emphasis on the repeatability, ac- nect, is a dramatic increase in connec- space communication and sens-
curacy and reliability of mmWave tion repeatability when compared to ing. Each of these new applications
test systems necessitates more au- manual insertion and connection. needs a robust supply chain of in-
tomation and less manual effort in Though the repeatability of properly terconnect options and will benefit
moving mmWave devices and sys- torqued threaded coaxial connec- from more automated manufactur-
tems between calibration and vari- tors by a skilled technician may be ing and testing approaches.

18 5G/6G  MAY 2024

Website : us.tecdia.com
 FOCUS
SPECIALFOCUS TECHNICAL FEATURE

Paving the Way to Terabit Wireless

Telecommunications
Dr. John Howard, Rana Barsbai and Steve Jalil
Electromagnetics Technology Industries, Inc., Boonton, N.J.

WIRELESS TECHNOLOGY The net effect is to reduce the spec- sorption plus scattering caused by
TODAY tral efficiency, meaning the increase hydrometeors2,3 in the transmission

T
oday, data has become a in data throughput speed is sub- medium depolarizes the transmit-
major part of the digital stantially lower than the bandwidth ted radiation. This effect may se-
existence. The demand increase would imply. The noise verely limit system performance,
for faster, more reliable power is determined by the formula particularly in the case where two
and ubiquitous connectivity has in Equation 1. orthogonal polarizations are used
reached unprecedented heights. N=KTB (1) as separate communication chan-
The U.S., home to around 144,000 nels. Figure 1 shows a plot of the
to 145,000 telecom towers, stands Where: atmospheric attenuation of RF
at the forefront of this digital revolu- N = Noise power (W) signals and how this attenuation
tion. These towers have long been K = Boltzmann’s constant = 1.381 × changes with frequency for various
the backbone of communication 10-23 W/Hz/K rainfall rates. This shows the effect
networks, facilitating both fixed and T = 290 K at room temperature of absorption and scattering on RF
mobile communications. However, B = RF carrier bandwidth (Hz) signals in varying rain conditions
now the FCC proposes the national and the substantial increase in loss
fixed broadband standard requir- Incorporate mmWaves per km for the mmWave frequency
ing 100 Mbps for download speeds Frequencies from 30 to 300 GHz range is clear.
and 20 Mbps for upload speeds. are considered mmWaves. In 5G
telecommunications technology, Increase Modulation Indices
To support that standard and to
mmWaves are employed to provide Higher modulation indices, like
continue to enable the demands
high data rates and low latency in 64-QAM, 256-QAM or 1024-QAM,
of ubiquitous connectivity, data
throughput speeds must increase. wireless communication systems.
There are various methods to in- However, mmWave signals have 100
Rainfall Rate = 150 mm/hr
crease wireless telecommunications propagation limitations; they are
data throughput speeds and this more susceptible to atmospheric 100
section will address several of these absorption and scattering1 and 50
Attenuation (dB/km)

techniques. obstacles like buildings and foli- 10 25

age can disrupt propagation. To 12.5
More Channel Bandwidth compensate for these limitations, 5
It is well known that as the chan- networks typically use higher trans- 2.5
nel bandwidth increases, the data mission powers and this introduces 1
throughput speeds in bits per sec- more noise. In these noisy channel
ond also increase. Unfortunately, an environments, whether the noise is Temp = 20° C
increase in bandwidth also increas- caused by atmospheric absorption,
es the effects of noise. For a con- scattering, reflections or if signal
0.1
stant power output, this increase in sidelobes are interfering, this noise 1 10 100 300
noise acts to reduce the signal-to- will impair the ability to increase Frequency (GHz)
noise ratio (SNR). This, in turn, re- data throughput speeds.
duces the modulation index, which Adding to the challenges, ab-  Fig. 1 RF signal atmospheric
reduces the data throughput speed. attenuation versus frequency.

20 5G/6G  MAY 2024

 FOCUS
SPECIALFOCUS TECHNICAL
COVER FEATURE

increase data throughput speeds of each beam of the communication this four-beam MIMO beamform-
but these modulation indices re- link. This reduction in antenna gain ing network. The system uses a 20
quire high SNRs to reap a substantial reduces the SNR and this will reduce MHz channel in a 120-degree sector
benefit from the increase in modula- the modulation index and decrease to provide an average of 28 Mbps
tion complexity.4 Table 1 shows an the data throughput speeds. These data speed in an urban environment.
example of SNR requirements for efforts can quickly create a law of By implementing the four-beam ar-
different modulation schemes and diminishing returns situation. Con- chitecture shown in Figure 4 in the
coding rates. In addition to the at- ceptually, the gain reduction expe- same 120-degree sector and repeat-
mospheric attenuation of the RF rienced with multiple beams from a ing the same 20 MHz bandwidth for
signal, there will also be spreading MU-MIMO antenna system is shown each beam, data speeds increased
loss with distance that will depend in Figure 2. by a factor of 10.6 The results of the
on the geometry of the transmit/ existing network and the ETI network
receive antenna array. The effect of A SOLUTION are presented in Figure 5.
these losses will be to reduce the The four methods to increase Earlier, the article discussed the
received signal while the interfering data throughput are all currently signal propagation challenges for
noise increases, reducing the SNR. in use and enable higher data signals in the mmWave frequency
This will decrease the modulation throughput rates. However, as dem- range. The reason for going to
index of the system, reducing the onstrated, there are challenges as- mmWave frequencies was to access
data throughput speeds. sociated with each method. This wider bandwidth channels to obtain
section provides an alternative solu- higher data rates, but the band-
Implement MU-MIMO width reuse architecture obviates
tion that avoids or minimizes these
Multi-user MIMO (MU-MIMO) challenges. The previous analysis this need. The ETI system addresses
is like troposcatter diversity. This showed that even though an in- the signal propagation issues by us-
technique permits separate data crease in bandwidth increases the ing frequencies below 10 GHz. This
streams to propagate in parallel. Us- throughput speeds, it also increases earlier section also discussed how
ing multiple data streams increases the noise. As a result, the modula- absorption plus scattering could
the data throughput speeds with tion index decreases to keep the depolarize orthogonal transmission
the increase dependent on channel link closed and this reduced spec- signals. U.S. patent 10,141,640 B2
conditions. tral efficiency partially negates the
Creating multiple antenna beams benefit of larger bandwidth for data 1 Beam
to transmit multiple data streams throughput speeds.
means that only a portion of the full 2 Beams
The alternative solution that is be-
phased array antenna is used for ing developed maintains the same 4 Beams
each beam.5 This reduces the EIRP bandwidth in a sector but repeats
this bandwidth
TABLE 1 multiple times
SNR REQUIREMENTS FOR DIFFERENT within the sector.
MODULATION SCHEMES AND CODING RATES The bandwidth  Fig. 2 Gain reduction in MU-MIMO
system.
REQUIRED BASEBAND SNR
reuse scheme be-
SNR REQUIREMENTS VS. CODING RATE AND MODULATION SCHEME haves similarly
Modulation Code Rate SNR (dB)
to an increase in
bandwidth, but it
1/8 -5.1
does not increase
1/5 -2.9 the noise of each of
1/4 -1.7 the beams. The net
1/3 -1.0
effect is higher data
QPSK throughput speeds
1/2 2.0 and higher spectral
 Fig. 3 Rendering of a simple four-
beam azimuth pattern.
2/3 4.3 efficiency. Figure 3
3/4 5.5 shows a rendering A1 A3 A2 A4

4/5 6.2
of the four-beam
radiation pattern in 90° Hybrid 90° Hybrid
1/2 7.9 azimuth and Figure 45° 45°
2/3 11.3 4 shows the simple
16-QAM
3/4 12.2 architecture of this
4/5 12.8
MIMO beamform- 90° Hybrid 90° Hybrid
ing network.
2/3 15.3 Electromagnet- M1 M3 M2 M4
64-QAM 3/4 17.5 ics Technology In-
4/5 18.6 dustries (ETI) has  Fig. 4Four-beam MIMO
built and deployed beamforming network.

5G/6G  MAY
5G/6G  2024
MAY 2024 21
21
 FOCUS
SPECIALFOCUS TECHNICAL
COVER FEATURE

discusses MIMO Figure 7 shows the results of im-

350
Competitor ETI dual-polarization plementing these proprietary and
300 307 306 303
scenarios and patented techniques. From the re-
300
250
Mbps Mbps Mbps Mbps these ideas are sults, we see that the sidelobe levels
200
fundamental to the have been reduced to more than 38
150
ETI architecture.7 dB below the main lobe with con-
Higher modula- stant beamwidth. These sidelobe
100
tion indices were levels remain roughly constant as
50
27 Mbps 28 Mbps 28 Mbps 28 Mbps shown to increase the antenna is scanned along its azi-
0 Week 1 Week 2 Week 3 Week 4 data throughput muth angle.
speeds, but this Figure 8 shows a second tech-
 Fig. 5 ETI four-beam system speeds versus the original increase came with nique that reduces sidelobe levels
network speeds.
the requirement of even more. In this case, the sid-
ETI Multibeam Antenna System 23-beam ∼90-Degree Azimuth higher system SNR elobes are more than 50 dB below
0 values. Achieving the main lobe. Using this technique
these higher SNRs maintains this sidelobe reduction
requires higher performance over an even broader
gain antennas and azimuth scan angle. The results of
–10
multiple, narrow Figure 7 and Figure 8 raise the pos-
beams. Passive and sibility of using these techniques
active beamform- to improve antenna performance
ing networks in azi- enough to increase the modulation
Gain

–20
muth and elevation index and improve data throughput
can easily enable rates.
the antenna gain
–30 and beam char- TERABIT WIRELESS SOLUTIONS
acteristics neces- Enhancing the performance of
sary to support the wireless networks involves increas-
–40 higher modulation ing the capacity and data through-
0 50 100 150 indices. put speeds provided by wireless
Degrees Finally, the MU- base station antennas. This can be
MIMO section pre- done with multiple radiating lobes
 Fig. 6 23-beam ETI antenna system with 90-degree azimuth sented a method in both azimuth and elevation to
coverage.
to increase data effectively cover a designated geo-
throughput speeds graphic volume. Using the propri-
with multiple bit streams transmit- etary and patented sidelobe reduc-
0
ted by an antenna array capable of tion techniques described earlier,
–10 generating multiple simultaneous along with the antenna architecture
beams. When the antenna array that the article describes, allows for
Gain (dB)

–20 generates multiple beams, the gain modulation indices of 256-QAM

–30
of each beam is reduced from the and 1024-QAM. These techniques,
gain that would result if all the an- along with repeating the full band-
–40 tenna array radiators were used to width in each beam, can help
produce a single beam. To increase achieve higher capacity and data
–50
0 50 100 150 antenna beam gain, beam shap- throughput speeds. Figure 9 shows
Azimuth (°) ing networks can be added to the a rendering of the radiation pattern
beamforming networks mentioned
 Fig. 7 Sidelobe reduction previously.
techniques.
Figure 6 shows a gain plot for a
0
23-beam ETI antenna system in a
90-degree azimuth sector. As the
–10 plot shows, the sidelobe levels are
more than 25 dB from the main lobe
Gain (dB)

–20

–30
as it is steered in its azimuth angle.
Figure 6 also shows the gain reduc-
–40 tion at the edges of the 90-degree
–50 scan from broadside. This gain re-
0 50 100 150
duction as the beam is scanned
Azimuth (°)
from broadside can be corrected
by proprietary and patented tech-  Fig. 9 Phased array with multiple
 Fig. 8 Improved sidelobe reduction niques.
beams in azimuth and elevation. Source:
courtesy of DARPA.
technique.

22 5G/6G
5G/6G
MAY
MAY2024
2024
 FOCUS
SPECIALFOCUS TECHNICAL
COVER FEATURE

of an antenna system that employs six beams in elevation. architecture are interesting. A net-
beamforming in the azimuth and el- Table 2 shows the maximum work using this architecture with the
evation directions. speeds available for various channel 32-radio configuration, 1024-QAM
sizes using 1024-QAM modulation modulation and 80 MHz channel
Examples of the Technology and Table 3 shows the same data bandwidth will provide an aggre-
Traditional macrocell tower de- for 256-QAM modulation. gate speed of 18.4 Tbps from only
ployments use three 120-degree The number of radios per anten- 120 telecom towers. To put this into
sectors. This first example uses na can change based on the num- perspective, a geographical area of
three antennas and each antenna ber of beams that are required. In 348 square miles, roughly the size of
has 24 radios. Figure 10 shows a the following example, the network Lehigh County in Pennsylvania, can
drawing of the tower that supports again has three antennas, but in this distribute 128.4 Mbps to 143,000
the antenna and radio infrastructure case, each antenna has 32 radios. households from these 120 towers.
at a wireless base station. The deployment is the same as This is without factoring in a conten-
Figure 11 shows the concept of shown in Figure 11, but the radios tion ratio, but this service still fulfills
the radios housed inside a tower provide four beams in azimuth and the U.S. requirements for broad-
for a three-antenna system. For the eight beams in elevation. band.
data results that follow, the radios Table 4 shows the maximum
generate four beams in azimuth and speeds available for various channel CONCLUSION
sizes using 1024-QAM modulation This article has compared various
in this new radio configuration and methods currently used through-
Table 5 shows the same data for out the industry to increase data
256-QAM modulation. throughput speeds. Each one of
The practical implications of the these methods has advantages and
disadvantages. To minimize the dis-
advantages of these techniques, the
article has also presented a simple
solution based on a cascadable ra-
dio architecture that uses passive
beamforming networks, along with
proprietary and patented antenna
sidelobe techniques. Results show
that data throughput speeds, even
in simple cases, can increase sub-
stantially. The benefit of this tech-
nique is it enables operators to
meet emerging broadband require-
ments more readily. In the words of
 Fig. 10 Mounting configuration for  Fig. 11 Radios housed inside the Leonardo da Vinci, “Simplicity is the
three-sector base station. tower.
ultimate sophistication.”
TABLE 2 TABLE 4 References
MAXIMUM SPEEDS FOR MAXIMUM SPEEDS FOR 1. H.C. Van de Hulst, “Light-scattering by Small
1024-QAM MODULATION, 1024-QAM MODULATION, Particles,” New York: Wiley, 1957, pp. 28–36.
24-RADIO CONFIGURATION 32-RADIO CONFIGURATION 2. J. Howard and N.A. Mathews, “Crosspolarisa-
tion of Microwaves due to Rain on a Satellite to
Channel Size Total Maximum Channel Size Total Maximum Earth Path,” IEEE Trans. Antennas and Propa-
(MHz) Speed (Gbps) (MHz) Speed (Gbps) gate, Vol. AP 27, No. 6, November 1979, pp.
80 115 80 153 890–891.
3. G. Brussaard, “A Meteorological Model for
40 57 40 76 Rain-Induced Crosspolarisation,” IEEE Trans.
20 28 20 38 Antennas and Propagate, Vol. AP 24, No. 1,
January 1979, pp. 5–11.
4. Dr. O. Werther and R. Minihold, “LTE: System
TABLE 3 TABLE 5 Specifications and Their Impact on RF & Base
Band Circuits,” Rohde & Schwarz Application
MAXIMUM SPEEDS FOR MAXIMUM SPEEDS FOR Note, 2013.
256-QAM MODULATION, 256-QAM MODULATION, 5. E. Björnson, “Basics of Antennas and
24-RADIO CONFIGURATION 32-RADIO CONFIGURATION Beamforming-Massive MIMO Networks,”
2018, Web: https://www.youtube.com/
Channel Size Total Maximum Channel Size Total Maximum
watch?v=xGkyZw98Tug.
(MHz) Speed (Gbps) (MHz) Speed (Gbps)
6. J. Howard, “World’s First Wireless Fiber,” The
80 92 80 122 North Jersey IEEE MTT/AP Societies 37th Annu-
al Symposium and Mini-Show, October 4, 2023.
40 46 40 61
7. J. Howard, “Isolation of Polarizations in Multi-
20 23 20 30 Polarized Scanning Phased Array Antennas,”
US 10, 141, 640 B2, 27, November 2018.

24 5G/6G  MAY 2024

M M W A V E F I LT E R S

LTCC Meets 5G
The World’s Widest Selection

• Band pass filters optimized for n257, n258, n260

and n261 5G bands
• Low pass filters with passbands up to 32 GHz
• High pass filters with fco up to 57 GHz
• Rejection up to 100+ dB
• Proprietary material systems and distributed topologies
• Pick-and-place standard case styles
 FOCUS
SPECIALFOCUS PRODUCT FEATURE

Python Package Controls AWG DDS

Multi-Tone Generation
Spectrum Instrumentation
Grosshansdorf, Germany

S
pectrum Instrumenta- tors (AWGs). The AWGs offer output ing the abstraction layer of the new
tion released a new rates of up to 1.25 GSPS and band- Python package makes program-
direct digital synthesis widths of up to 400 MHz. The op- ming the AWG hardware and the
(DDS) firmware option tion allows users to define 23 DDS DDS mode extremely easy. A DDS
for the company’s range of versatile cores per AWG that can be routed example with 16 superimposed sine
16-bit arbitrary waveform genera- to the hardware output channels. waves on one channel and the FFT
Each DDS core can is shown in Figure 1.
be programmed
for frequency, am- DIRECT DIGITAL SYNTHESIS
plitude, phase, DDS is a method for generat-
frequency slope ing arbitrary periodic waves from
and amplitude a single, fixed-frequency reference
slope. This enables clock. It is a technique widely used
control of lasers in a variety of signal-generation
through acousto- applications. The DDS functionality
optic deflectors implemented on Spectrum’s AWGs
and acousto-optic is based on the principle of adding
modulators, as is multiple DDS cores to generate a
often used in quan- multi-carrier (multi-tone) signal,
tum experiments, with each carrier having a well-
with just a few sim- defined frequency, amplitude and
ple commands. Us- phase.
 Fig. 1 16 superimposed sine waves and the FFT.

26 5G/6G  MAY 2024

 Unlocking the Power of Waveguide

GUARANTEED PERFORMANCE FROM 1GHz TO 1.1THz

MOTORISED LINEAR ATTENUATOR

DEMONSTRATED AT IMS 2024

We are pleased to expand our offering

in the growing automated test
equipment category. Our new, cost
effective, compact variable attenuator
boasts an impressive: 0-50 dB range
& ± 10% attenuation flatness, across
the whole attenuation band.
COMPACT ANTENNA The team have setup a demonstration on booth 1152
RANGE where you can see our new motorised attenuator in
action. For technical details see the Product Brief in
Flann have a complete range this publication.
of waveguide instruments to
fulfil your needs up to 330 GHz.
Our instruments, including our
unique OMT, enable you to
perform full-band validation
testing..

GET
IN TOUCH:

flann.com
sales@flann.com
OUR EXPERIENCE AND CAPABILITY PROVIDES RELIABLE,
+44 (0)1208 77777
POWERFUL, HIGH PERFORMANCE SOLUTIONS
@flann-microwave
 FOCUS
SPECIALFOCUS PRODUCT
COVER FEATURE

Card Trigger phase and frequency modulation

Sequence 1 2 3 on a carrier.
Command Queue Output

dds[0].freq(20e6)
(To Scale)
120 MHz
PYTHON PACKAGE FOR EASY
dds[0].amp(1.0) Core 0
+10 MHz/s
CONTROL OF DDS
(a) dds.trg_src(spcm.SPCM_DDS_TRG_SRC_TIMER) Frequency
dds.trg_timer(5.0) 20 MHz The Python package is avail-
dds.exec_at_trg()
Core 0 1.0 1.0 able through GitHub with a single
dds[0].freq(20e6)
dds[0].amp(1.0)
Amplitude 0.6
pip command. It allows full control
(b) dds.trg_src(spcm.SPCM_DDS_TRG_SRC_TIMER) Timer 5s 10 s of all current Spectrum hardware,
dds.trg_timer(5.0) Trigger
dds.exec_at_trg() 0 5 15 Time (s)
including digitizers, AWGs, digital
dds[0].frequency_slope(0) I/O cards and options, including the
(c)
dds[0].amp(0.6) DDS option. Generating a single
dds.trg_src(spcm.SPCM_DDS_TRG_SRC_NONE)
dds.exec_at_trg() 100 MHz sine wave requires five
dds.write_to_card() DDS-specific single-line commands
as follows:
 Fig. 2 DDS command block and response. dds.reset()
dds[0].amp(0.5) # 50% output am-
plitude
dds[0].freq(100e6) # 100 MHz signal
frequency
dds.exec_at_trg()
dds.write_to_card()
In total, there are more than 10
different core-related functions, as
well as more than 30 general func-
tions, realized inside the Python
DDS class. This allows users to read
all the internal parameters and read
back all the possible DDS settings.
10 Million DDS Commands Per
Second
The DDS commands can be sent
using an extremely fast DMA mode
into the 4 GB sample memory of the
AWG. This fast transfer mode allows
 Fig. 3 DDS command script. more advanced DDS functions that
are not implemented in firmware as
DDS Mode Application Examples slope functions to produce extreme- intrinsic functions to be performed.
For years, Spectrum AWGs have ly smooth changes to frequency and These include:
been successfully used worldwide amplitude. Only a few Python com- • S-shaped frequency/amplitude
in pioneering quantum research ex- mands are necessary to generate a slopes consisting of multiple lin-
periments. The flexibility and fast sine wave, shown in the top block on ear slope commands.
streaming mode of the AWGs en- the left of Figure 2, ramp up the fre- • Custom frequency/amplitude
ables data to be streamed straight quency, shown in the middle block, slopes consisting of multiple lin-
from a GPU and allows the control and reduce the amplitude, shown in ear slope commands
of qubits directly from a PC. While the bottom block. • AM modulation consisting of
using an AWG in this way offers full The DDS option provides an multiple amplitude change com-
control of the generated waveforms, easy and programmable way for us- mands
large amounts of data need to be ers to produce trains of waveforms, • FM modulation consisting of
calculated. This slows the critical frequency sweeps or finely-tunable multiple frequency change com-
decision-making loop. In contrast, references of various frequencies mands
using the versatile multi-tone DDS and profiles. Applications that re- • FSK modulation consisting of
functionality greatly reduces the quire the fast frequency switching multiple frequency change com-
amount of data that must be trans- and fine frequency tuning that DDS mands.
ferred, while still maintaining full offers are widespread. They can A simple s-shaped frequency
control. All the key functionality re- be found in industrial, medical and slope can be achieved by using mul-
quired for quantum research is built imaging systems, network analysis tiple intrinsic linear frequency slope
in. With just a single command, us- or even communication technol- functions. The example in Figure 3
ers can apply intrinsic dynamic linear ogy, where data is encoded using defines a command timer of 100 ms

28 5G/6G  MAY 2024

High Resolution
Broadband
Multi-Octave
Phase Shifters Designed for Industrial & Military Applications
for mission-critical applications including Beamforming Modules, Frequency
Translation & Nulling, Beam Steering,
TR Modules, Communications Antennas,
Phased-Array Radar, Phase Modulators,
& Test Instrumentation
0 to 360°, High Resolution
Broadband & Multi-Octave up to 40 GHz
Digital & Analog Control available
Fast Speed, Low Insertion loss,
High Phase Accuracy
Hermetic Sealing, Military or Aerospace
Screening available
Form, Fit, Function & Custom Design Services

Quantic PMI offers a full portfolio of RF & Microwave Solid State Phase Shifters
available in high resolution, multi-octave & broadband designs up to 40 GHz.

Featured Products PS-200M10G-8B-SFF PS-812-360-14B-SFF PS-360-DC-3 OPTION PS-2G18G-360-12D-TS

Digital Phase Shifter OPT10D27 618-15D-10BIT Vector Phase Shifter
Digital Phase Shifter Digital Phase Shifter

Frequency Range 0.2 to 10 GHz 8 to 12 GHz 6 to 18 GHz 2 to 18 GHz

Phase Shift Range 360° 360° 360° 360°

Phase Accuracy ±0.5° Typ ±3° Max ±8° (Peak to Peak) ±15° Typ

Insertion Loss 8 dB 12 dB Typ 12 dB Typ 18 dB Max

Switching Speed 250 ns 100 ns 30 ns 500 ns Max

Control Configuration 8-Bit TTL 14-Bit TTL 10-Bit TTL 12-Bit TTL

Dimensions (inches) 3.25" X 3.25" X 0.84" 3.25” x 2.00” x 0.50” 1.60" x 1.75" x 0.50" 4.25" x 3.50" x 1.00"

Connectors SMA (F) SMA (F) SMA (F) SMA (F)

Get a complete list of our solid state phase shifters models

and available options, download product features and other
Phase Shifter Product technical data at quanticpmi.com

Or contact us at 1-877-752-6271 or sales@quanticpmi.com Don’t miss us.

Washington DC, June 16 - 21
© Quantic PMI (Planar Monolithics) 2024
Booth 1251
 FOCUS
SPECIALFOCUS PRODUCT
COVER FEATURE

chased 66xx-series products can be equipped with the

new firmware option by simply performing a firmware
update. Besides the high-level Python package, pro-
gramming can be done using the existing driver SDKs
that are included in the delivery. Examples are available
for C, C++, JAVA, C#, MATLAB, LabVIEW and many
more.
One Python Package for All Spectrum Products
The high-level, object-oriented Python package not
only supports the DDS mode, but it also supports all
current Spectrum products and nearly all product fea-
tures and operating modes. This includes digitizer,
AWG, synchronization, single-shot mode, continuous
FIFO acquisition, averaging modes, pulse generator
and DDS options, along with many more. More than 30
examples and detailed documentation are available on
 Fig. 4 Examples of the 66xx AWG series. how to use the package. Spectrum is constantly improv-
ing Python support and continues to add new examples
to generate an s-form slope in seven steps where each
and new features. All 23 of the 66xx AWG series, some
step is 100 ms. The slope goes from 100 to 122 MHz.
of which are shown in Figure 4, can be used with the
Availability and Software Integration DDS option.
The DDS option is available for the full range of
M4i.66xx PCIe cards, M4x.66xx PXIe modules, portable Spectrum Instrumentation
LXI/Ethernet DN2.66x units and multi-channel desktop Grosshansdorf, Germany
LXI/Ethernet DN6.66xx products. All previously pur- www.spectrum-instrumentation.com

30 5G/6G  MAY 2024

 FOCUS
SPECIALFOCUS PRODUCT FEATURE

MMIC Filter Design Goes Mainstream

with Web-Based Design Tools
Marki Microwave
Morgan Hill, Calif.

S
ize, weight and power must be fast-turn designs that are This makes them well-suited for
(SWaP) are key consid- right the first time. adaptive filtering applications such
erations in the compo- In recent years, Marki Microwave as LO tracking. These impedance-
nent selection process has developed a novel design flow insensitive filters utilize a balanced
for RF systems. MMIC processes for MMIC filter design including design to feature low return loss in
have a proven track record as the standard response types (band- both the passband and stopband.
technology platform for high per- pass, highpass and lowpass) as well Table 1 shows the performance re-
formance components throughout as some more advanced concepts sults.
the various sub-blocks in the RF like notch, reflectionless and con- Figure 1 demonstrates the tun-
signal chain. However, MMIC tech- figurable filters. Marki Microwave’s ing capability of Marki Microwave’s
nology has become a viable option proprietary design flow has dem- MFBT-00003PSM, which features a
for high volume filter development onstrated a high level of agree- wide center frequency tuning range
only recently. This is largely due to ment between initial simulations of 8 to 30 GHz. The top chart in Fig-
the growing importance of size re- and measured results, leading to ure 1 shows the MFBT-00003PSM
duction. first-pass design success for custom sweeping lowpass and the bottom
In an ideal world, filters would filter solutions. This enables rapid chart shows the MFBT-00003PSM
be selected early in the system filter development and reduced sweeping highpass.
design phase. These filters can be cycle times. Due to the inherent
commercial off-the-shelf solutions advantages of the MMIC platform, FILTER DESIGN TOOLS &
and in some cases, frequency plans these solutions are repeatable from PRODIGY™ FILTER DESIGNER
can be changed to accommodate unit to unit and wafer to wafer and Figure 2 shows several filter de-
filters available in the market. Un- are scalable to high volume. sign tools that Marki Microwave
fortunately, due to the presence of hosts on its website to enable the
unexpected spurious products that RECENT MMIC FILTER development of accurate fast-turn
may only present themselves dur- DEVELOPMENTS solutions:
ing testing, a real-world receiver MMIC is well suited not only for LC Filter Design Tool: Uses a
will also have several “oh no” fil- common filter response types but lumped element model to calculate
ters. The job of these filters is to also for more advanced concepts. LC filter circuit values for lowpass,
suppress these unexpected tones. Marki Microwave recently released a highpass and bandpass responses,
For projects where time to market family of varactor diode tunable fil- using either Chebyshev, Elliptic,
is key, this presents a problem as ters that allow users to create band- Butterworth, Bessel or Legendre fil-
these filters will be custom designs pass filters with variable center fre- ter topologies. It is a great tool to
that require engineering and manu- quencies and percent bandwidths see whether a filter is theoretically
facturing lead-times. For this rea- through independent analog tuning possible. If a filter cannot be real-
son, modern custom filter designs of the highpass and lowpass sides. ized using this tool, it is unlikely that
it is a realizable filter.
TABLE 1 Microstrip Filter Tool: Provides
MMIC TUNABLE FILTER RESULTS an ideal design of a distributed-
Center 3 dBc Tuning
element microstrip filter. This tool
Package allows you to design on various
Frequency, fc Passband Voltage
dielectric substrates, using Cheby-
MFBT-00001PSM 3.5 to 9.5 GHz 3 to 10 GHz 0 to 16 V 4x4 mm QFN
shev or Butterworth filter types, to
MFBT-00002PSM 5.5 to 15.5 GHz 4.5 to 16.5 GHz 0 to 16 V 4x4 mm QFN get a first approximation of a filter
MFBT-00003PSM 10 to 26 GHz 8 to 30 GHz 0 to 16 V 4x4 mm QFN design. This demonstrates what is

32 5G/6G  MAY 2024

Technology. Support. Success.
Powerful EM Simulation Software and Real-World Expertise.

Electromagnetic
Simulation Software

XFdtd® 3D EM Simulation Software has evolved alongside modern antenna design, meeting
industry requirements and solving today’s most advanced antenna design challenges.

XFdtd's new schematic

editor enables advanced
antenna matching network
and corporate feed network
analyses as part of a
unified solver workflow,
including multi-state and
multi-port devices.

Explore all of XFdtd’s powerful features at www.remcom.com/xfdtd

Superior Support
Remcom provides an extraordinary level of support to our customers, giving you direct
access to EM experts who are dedicated to your success. Our team is there to guide
you through all of your EM challenges, from the simple to the technically complex.

Learn more about the Remcom difference at www.remcom.com/about-remcom

+1.888.7.REMCOM (US/CAN) | +1.814.861.1299 | www.remcom.com

 FOCUS
SPECIALFOCUS PRODUCT
COVER FEATURE

possible, but results should be verified and optimized WHAT IS

in 3D EM simulations. PRODIGY?
Prodigy Filter Designer: Custom filter solutions of- The LC Filter
ten need to be developed quickly to prevent delays. Design Tool and

Passband Width
The standard filter development flow is an iterative Microstrip Filter De-
process between a customer and manufacturer on filter sign Tools can be
specification negotiations and system validation that thought of as calcu-
can take weeks. Marki Microwave developed Prodigy lators to see what
Filter Designer to streamline the development flow for is theoretically pos-
system designers and reduce time to market. sible, whereas Prod-
igy Filter Designer is
a MMIC filter design
0 tool that produces a
real FEM-designed 
Fig. 3 Matrix of known good filter
–10 designs generated by HOTMESS.
filter with known
–20
design variables and size. It uses machine learning to
–30 calculate real S-parameters nearly instantaneously,
–40 including all the 3D effects like metal loss, parasitics,
0 V/16 V cross-coupling, etc. Prodigy is powered by HFSS opti-
–50
4 V/16 V mization through microwave element surrogate simu-
8 V/16 V
–60
12 V/16 V lation (HOTMESS), which is Marki Microwave’s propri-
–70 16 V/16 V etary automated 3D FEM solver. HOTMESS generates
–80
a set of 3D FEM-validated S-parameters, including
0 5 10 15 20 25 30 35 40 the physical layout of the filter. Marki Microwave uses
Frequency (GHz) HOTMESS to create databases of known good filters
by filter type and filter topology. This is the foundation
0
of the Prodigy Filter Designer. Prodigy takes the known
–10 good filter designs and uses machine learning algo-
–20 rithms to interpolate between these points to design
–30
a filter based on customer inputs. This idea is shown,
conceptually, in Figure 3.
–40
Users can select the filter best suited to their specific
0 V/0 V
–50
0 V/4 V
requirements and input desired center frequency and
–60 0 V/8 V percent bandwidth from the growing list of topologies.
0 V/12 V The S-parameter file and die dimensions of their filter
–70 0 V/16 V
are instantly provided from these inputs. Designers can
–80
0 5 10 15 20 25 30 35 40 then take that S-parameter file and validate the design
Frequency (GHz) in their system. Prodigy essentially pushes the iterative
process of specification negotiation to the customer
 Fig. 1 MFBT-00003PSM insertion loss. without having to involve a Marki engineer, saving
weeks in the process. Once they are happy with a de-
sign, Marki’s engineers take it from
there. If a customer desires a filter to
LC Filter Design Tool be packaged into a surface-mount
Fundamental Filter Design, package, Marki’s engineers can take
Provides the Theoretical
Performance of Various Filters LC Filter Design Tool their filter design, apply packaging
effects and send back a validated set
Microstrip Filter Design Tool
of S-parameters.
Simple Physical Representation
of a Planar Filter, Provides Currently, Prodigy can be used to
Microstrip Filter
Insight into What is Possible design bandpass filters and it will con-
Design Tool
NEW
tinue to grow as more topologies and
HOTMESS filter response types are added. Prodi-
A Propriety Automated Filter gy Filter Designer is currently available
NEW
Solver that Leverages 3D
3D FEM & Machine Prodigy Filter Designer
on www.markimicrowave.com.
FEM and Machine Learning
Learning Filter Solver A Propriety Automated Filter
Algorithms
Design Tool with Excellent
Correlation Between Simulated
Marki Microwave
and Measured Results Morgan Hill, Calif.
www.markimicrowave.com

 Fig. 2 Marki Microwave’s web-based filter design tools.

34 5G/6G  MAY 2024

 11:48
W
1:03

10:05
fifififi

9:00
3:14
A

fififififififi

4:09
fififififififi

5:00
N
v
v

ll
®

W
“Y

h
w h
L KF

.lp fu .c m/pcb
1-800-345-L KF
 FOCUS
SPECIALFOCUS TECH BRIEF

Bootstrap Oscillator Delivers

Exceptional Phase Noise Performance
T
he Quantic Wenzel boot- and 130 MHz, the oscillator pro- days of continuous operation.
strap oscillator is a cut- vides versatility for diverse appli- • Power Supply Stability: An in-
ting-edge frequency con- cations. ternal voltage regulator enhanc-
trol and timing solution • Phase Noise Performance: With es power supply line rejection,
designed for applications requiring a static phase noise of -188 dBc/ contributing to stable and reli-
ultra-low phase noise performance Hz and dynamic phase noise of able operation.
and low g-sensitivity in harsh environ- -165 dBc/Hz, the oscillator ensures Established in 1978, Quantic
ments. Operating between 80 MHz precision in signal generation. Wenzel is a pioneer in ultra-low
and 130 MHz, the bootstrap oscilla- • Vibration Compensation: Offered phase frequency control and tim-
tor uses two phase-locked oven-con- in active and passive versions, the ing solutions. The portfolio ex-
trolled crystal oscillators (OCXOs) of oscillator exhibits a sensitivity of 5 tends to advanced frequency
the same frequency. The innovative x 10-11/g below 300 Hz offsets and sources and integrated microwave
technology cancels vibration-induced 5 x 10-12/g above 300 Hz offsets assemblies reaching 30 GHz and
phase noise by mechanically and with vibration isolation. beyond. These products find ap-
electrically aligning the oscillators. • Size and Housing: Housed in plications in mission-critical mili-
The result is a highly stable oscilla- a 5.86 x 3.7 x 1.18 in. machined tary, space and commercial envi-
tor assembly with exceptional low g- case, the oscillator meets space ronments, delivering reliable per-
sensitivity. constraints in airborne, mobile formance in extreme conditions.
and shipboard environments.
PERFORMANCE • Aging Stability: The bootstrap Quantic Wenzel
CHARACTERISTICS oscillator maintains reliability over Austin, Texas
• Frequency Range: Covering any time, with a typical aging rate of quanticwenzel.com
fixed frequency between 80 MHz ±1 x 10-6 in the first year after 30

RF
RF
from JFW
JFW

Test Systems Programmable Attenuators

Terminations Fixed Attenuators
Variable Attenuators RF Switches
Power Dividers RF Test Accessories

JFW
Ca 317-887-1340 T F 877-887-4JFW (4539)
E-ma a @jfw .c m V www.jfw .c m

36 5G/6G  MAY 2024

 FOCUS
SPECIALFOCUS NEW PRODUCTS

COMPLETE ANTENNA RACK-MOUNT ATTENUATOR SOLID-STATE SWITCHES

MEASUREMENT CHAMBER
SOLUTION Model 50PA-1019- Quantic PMI’s port-
XX is a 19 in. rack- folio of RF and mi-
Copper Mountain mount attenuator crowave, solid-state
Technologies’ has assembly contain- switches that range
partnered with Mil- ing 50 Ω solid-state up to 70 GHz. Fea-
liBox and Eravant to step attenuators. tures include ultra-
offer a complete The number of attenuators can be se- broadband perfor-
antenna measurement chamber solu- lected from 1 to 8 attenuators. The at- mance, ultra-high
tion. It is a turnkey system for customers tenuators have an attenuation range of isolation, low loss, ultra-high speed, ul-
needing to make far-field antenna mea- 0 to 63 dB by 1 dB steps and operate at tra-low power consumption, SPST -
surements above 18 GHz. The over-the- 100 MHz to 40 GHz. It is available with SP64T, reflective or absorptive avail-
air (OTA) antenna measurement chamber SMA female connectors. This attenua- able. Options include ultra-low video
solution features a wide array of configu- tor assembly can be remotely controlled transient models, high input power
rations depending on the user’s anechoic via an Ethernet connection or a serial models, custom package or form factor
chamber requirements, antenna size, de- connection. can be supplied (usually without a NRE
sired frequency range and antenna posi- charge), hermetic sealing, military or
JFW Industries space screening, amplitude and phase
tioner functionality. The result is an afford- www.jfwindustries.com
able and smaller than traditional chamber matching. Form, fit and function de-
testing solution. Learn more now! signs our specialty. COTS availability!
6G SUB-THZ REFERENCE Quantic PMI
Copper Mountain Technologies
www.coppermountaintech.com ARCHITECTURE www.quanticpmi.com
NI’s 6G Sub-THz
5G LOW NOISE AMPLIFIER Reference Architec- WIRELESS INSITE®
ture brings D-Band
measurement capa-
Experience un- The ability to con-
bility with PXI vec-
matched power trol the EM propa-
tor signal transceiv-
and precision with gation environment
ers and Virginia Di-
Fairview Micro- is an important area
odes (VDI) frequency extensions. With
wave’s compact, of 6G research.
support for Tx and Rx power, power and
coaxial 0.7 dB NF Remcom’s Wireless
flatness calibration, up to 4 GHz instan-
low noise amplifier, InSite® supports the simulation of RF
taneous bandwidth and real-time data
tailored to cover pivotal 5G frequency interactions with engineered electro-
streaming with an FPGA coprocessor,
bands. With an operating range from magnetic surfaces (EES), allowing mod-
this solution delivers a high-level start-
600 MHz to 4.2 GHz, our low noise am- eling of passive metasurfaces designed
ing point for configuring 6G sub-THz
plifier is a connectorized medium power to optimize wireless communication
research, prototyping and validation ap-
version with SMA connectors that is rated coverage by manipulating how signals
plications while enabling real-world
at 17 dBm, has a 21 dB typical gain and is propagate through a scene. Wireless In-
communication system prototyping and
available for same-day shipping. Find ad- Site’s ray-tracing and EM path process-
validation.
ditional specs and data sheets for ing calculations enable the prediction of
FMAM1079 at fairviewmicrowave.com. National Instruments (NI) reflections, transmissions and diffrac-
www.ni.com tions that interact with EES placed with-
Fairview Microwave in office buildings and urban areas.
www.fairviewmicrowave.com
NEX10 5G TERMINATION Remcom
OMT www.remcom.com
The OMT is a bi-di- Pasternack’s preci-
rectional device sion-engineered CERAMIC METALLIZATION
which allows signals terminations with a OPTIONS
to be sent from, or NEX10 connector- Remtec Inc. (Can-
received at, the ized design is craft- ton, Mass.) offers a
rectangular ports ed for seamless operation up to 6 GHz. newly expanded
via an antenna sup- With a 50 Ω impedance and a maximum range of metalized
porting dual polari- input power rating of 5 W, this termina- thick film ceramic
sation (at the common port). With models tion has an aluminum body and is de- substrate material
available from 33 to 220 GHz Flann’s full- signed to withstand temperatures be- options – including Alumina, BeO, BTO,
band OMTs offer the high performance tween -40°C and +75°C. Ideal for tele- Silicon Nitride and AIN. Combined with
demanded by today’s communication sys- communications, broadcast and advanced and proprietary screen print-
tems in which good polarisation purity and aerospace applications, these termina- ing, etching and plating capabilities –
isolation are paramount. These instru- tions ensure reliable signal modulation such as plated copper (PCTF®) and di-
ments have two rectangular ports on adja- and termination. Download data sheets rect bonded copper (DBC) -- Remtec is
cent faces and a common circular port. and shop for PE6TR1160 at our website. uniquely positioned to be the design
and production arm for customers re-
Flann Microwave Pasternack quiring single or multilayer ceramic cir-
www.flann.com www.pasternack.com. cuits for SMT modules, hybrid integrat-
38 5G/6G  MAY 2024
 FOCUS
SPECIALFOCUS NEW
COVER
PRODUCTS
FEATURE

ed circuits, integrated passives devices, ULTRA-LOW PHASE NOISE ACE9600 ADVANCED

antenna arrays, filters, comb generators, FREQUENCY SYNTHESIZERS CHANNEL EMULATOR
as well as optoelectronic and sensor ap- The KSSLO series of A major design
plications. Contact Remtec’s Business ultra-low noise fre- challenge for 5G
Development Manager to arrange a quency synthesizers non-terrestrial net-
free, thick-film capabilities presentation. offer unmatched works is the proper
Remtec Inc. phase noise perfor- evaluation of elec-
www.remtec.com mance in a variety tromagnetic and environmental effects
THERMAL POWER SENSOR of frequency bandwidths operating in fre- on the signal path, including noise,
quency range 1 and 2 allocations in 5G Doppler shift, latency and multi-path
and 6G NR systems and supporting test and fading within a laboratory environ-
Discover the R&S®NRP170TWG(N)
equipment applications. These products ment. The explosive growth of low earth
thermal power sensor, the market’s only
are user programmable for dynamic se- orbit satellites has intensified the need
NMI-traceable power sensor for precise
lection of desired TX and or RX channel for a flexible, COTS satellite link emula-
D-Band measurements, covering 110 to
bandwidths within the desired FR alloca- tion system such as the ACE9600 to
170 GHz. Engi-
tion, phase lock to the available external emulate the wireless path before satel-
neered for simplici-
master reference of the user system re- lite deployment. Visit booth 704 at
ty and top perfor-
quirement and have extremely low laten- IMS2024 for additional information.
mance, it integrates
cy due to the wide loop bandwidth, which Wireless Telecom Group
into any setup.
also allows for wideband frequency mod- www.wirelesstelecomgroup.com
Moreover, with fast,
ulation. Frequency selection is by three-
accurate and stable
wire SPI control. Phase noise within the
readings, as well as USB/LAN connectiv-
wide loop bandwidth is typically as low as
ity, it provides capabilities essential for
-116 dBc/Hz at 10 KHz and 100 kHz offset
6G R&D and advanced radar applica-
from the selected frequency when operat-
tions, allowing you to measure power
ing at 26 GHz. The product is housed in
with confidence and exceed expecta-
industry standard 1.25 × 1.25 × 0.5 in.
tions.
package. The operating temperature
Rohde & Schwarz GmbH & Co. KG
range is from -40°C to +65°C.
www.rohde-schwarz.com
Synergy
www.synergymwave.com

40 5G/6G  MAY 2024

7