Release: Multi Technology (3G and LTE) Whitepaper
Release: Multi Technology (3G and LTE) Whitepaper
073.05.01
Multi technology (3G and LTE)
Whitepaper
December 2013
www.scf.io/ www.smallcellforum.org
SMALL CELL FORUM
RELEASE Five
Four
We are not a standards organization but partner with organizations that inform
and determine standards development. We are a carrier-led organization. This
means our operator members establish requirements that drive the activities
and outputs of our technical groups.
Our track record speaks for itself: we have driven the standardization of key
elements of small cell technology including Iuh,FAPI/SCAPI, SON, the small cell
services API,TR‑069 evolution and the enhancement of the X2 interface.
At the time of writing, Small Cell Forum has more than 140 members, including
68 operators representing more than 3 billion mobile subscribers – 46 per
cent of the global total – as well as telecoms hardware and software vendors,
content providers and innovative start-ups.
This document forms part of Small Cell Forum’s Release Five: Rural &
Remote that considers the opportunities and perceived barriers associated
with the deployment of small cells in rural and remote scenarios, including
disaster recovery, military installations, as well as verticals such as oil and gas,
maritime, aviation and automotive.
The Small Cell Forum Release Program has now established business cases
and market drivers for all the main use cases, clarifying market needs and
addressing barriers to deployment for residential, enterprise and urban small
cells.
All content in this document including links and references are for informational
purposes only and is provided “as is” with no warranties whatsoever including
any warranty of merchantability, fitness for any particular purpose, or any
warranty otherwise arising out of any proposal, specification, or sample.
If you would like more information about Small Cell Forum or would
like to be included on our mailing list, please contact:
Email info@smallcellforum.org
This paper addresses the concept of combining two radio technologies 3G and 4G in
small cell networks by first highlighting the foreseen challenges in public space,
enterprise, and residential small cell deployments. Then exploring technology
synergies that can enable integrated solutions for multi-technology small cell
architectures, radio resource management systems, and traffic offload methods.
Four multi-technology small cell architectures are considered that operators can chose
from based on their legacy technology deployments and possible new technology
requirements:
Finally, traffic offload in multi-technology small cell networks is addressed for traffic
offload in the radio access networks and through the 3G and 4G gateways; and
benefits of traffic offload for operators and users are highlighted.
Tables
Table 4-1 Various Multi-RAT RRM models and associated use cases ....................17
Figures
Figure 3-1 Architecture A – Separate 3G and 4G networks .................................. 6
Figure 3-2 Architecture B – Combined management system ................................ 7
Figure 3-3 Architecture C-1 – Integrated security gateway (with separate IPsec
Tunnels for luh and S1 interfaces)..................................................... 8
Figure 3-4 Architecture C-2 – Integrated security gateway (with a combined
IPsec Tunnel for luh and S1 interfaces) .............................................. 8
Figure 3-5 Architecture D – Integrated HNB/HeNB gateway ................................. 9
Figure 4-1 Multi-RAT RRM in a multi-standard HetNet ........................................10
Figure 4-2 Multi-RAT RRM functionalities and procedures ...................................12
Figure 4-3 Multi-RAT RRM architecture .............................................................13
Figure 4-4 Interaction between multi-RAT RRM and local RRM entities .................15
Figure 4-5 Single access point – Integrated multi-RAT RRM topology ...................18
Figure 4-6 Multiple access point – integrated Multi-RRM Topologies, (a)
integrated multi-RAT RRM in each intra-RAT AP, (b) integrated multi-
RAT RRM in some APs, and (c) integrated intra-RAT in Single AP with
integrated multi-RAT RRM in some APs. ............................................19
Figure 4-7 Multiple access point – server/cloud-based multi-RAT RRM topology ....20
Figure 4-8 Integrated multi-RAT RRM: UE-based scenario ..................................20
Figure 4-9 Generic multi-radio access technology - RRM topology .......................21
Figure 4-10 Multi-tier hierarchical inter-operator RRM topology .............................21
Figure 5-1 Local data offload (LIPA) at multi-mode ESCC/ESCG ..........................24
Abbreviations
2G Second generation
3G Third generation
3GPP 3rd generation partnership project
C-plane Control plane
eNodeB Evolved node B
eNB E-UTRAN node B
EMS Element management system
EPC Evolved packet core
EPS Evolved packet system
EUTRAN Evolved universal terrestrial radio access network
GERAN GSM and EDGE radio access network
GPRS General packet radio service
GSM Global system for mobile communications
HeNB Home eNB
HeNB-GW HeNB gateway
HeMS HeNB management system
HNB Home NB
HNB-GW HNB gateway
HMS HNB management system
LAC Location area code
LTE Long term evolution
LIPA Local IP access
L-GW Local gateway
MME Mobility management entity
MSC Mobile switching centre
NMS Node B management system
OAM Operation and maintenance
PGW Public data network (PDN) gateway
RAC Radio admission control
RAN Radio access network
RRM Radio resource management
S-GW Serving gateway
S1 Interface between eNB, MME, and S-GW
S11 Interface between MME and S-GW
S1-MME S1 for the control plane
S1-U S1 for the user plane
SeGW Security gateway
SGSN Serving general packet radio serving support node
SIPTO Selected IP traffic offload
TAC Tracking area code
UE User equipment
UTRAN Universal terrestrial radio access network
U-plane User plane
WLAN Wireless LAN
1. Introduction
Global mobile data traffic is expected to increase many-fold over the coming years.
This skyrocketing demand is fuelled by the proliferation of data-hungry wireless
devices (e.g., smartphones, tablets, laptops), that are accessing mobile networks
worldwide.
Faced by this new reality, mobile operators will seek ways to meet the traffic demand
while avoiding any possible increase in capital expenses (CAPEX) and operational
expenses (OPEX) of their networks.
Data offload to small cells has become a popular choice of mobile operators to
improve the cellular network capacity in highly congested areas, or extend user
coverage to problematic cell-edge areas. More specifically, small cell deployments
have indicated enhanced user quality of services and a possible overall saving in
networks CAPEX and OPEX for operators.
Clearly, the way forward for mobile operators in the race of multi-technology market
will mean overcoming significant challenges to the multi-technology small cell
deployments related to:
The key objective of this paper is to explore technology synergies that can enable
integrated solutions for multi-technology small cell deployments. More precisely,
Section 2 of this paper describes the case for multi-technology small cell networks and
foreseen challenges in public, enterprise, and residential deployments, and proposes
ways to address such challenges. Section 3, explores alternative multi-technology
small cell architectures and their pros and cons in terms of optimization of product
integration and signaling. Section 4 presents technology synergies for combined radio
resource management and possible implementation models. Section 5, investigates
solutions of traffic offload for multi-technology small cell deployments. Finally, Section
6 presents the summary of whitepaper.
Public, enterprise, and residential access networks will continue to see a mix of both
LTE and 3G handsets for quite some time ahead; and although 4G handset market
penetration will increase, 3G handsets shall continue to represent a major fraction of
total mobile handset market.
In this light, there is a great need for the deployment of multi-technology small cell
networks that will be able to efficiently manage available operators’ network resources
between 4G and 3G subscribers. Take for example, the efficient use of available
spectrum by using the 4G spectrum for data services, and the 3G spectrum for voice
services.
Additional multi-technology small cells will benefit from the possibility to share the
backhaul power, and real estate, and deploy a common gateway, in order to connect
to the core network and leverage a unified management interface framework.
It is imperative that the networks should support increased data capacity and speeds,
while lowering latency and power consumption, all at significantly lower costs. More
specifically, operators seek ways to reduce network costs related to manufacturing,
power consumption and service delivery, while intelligently using available 3G, 4G
networks and Wi-Fi connectivity to support macro-cellular traffic offload and selection
of networks based on delivered services.
Future multi-technology small cell networks will face different challenges depending on
their deployment scenarios as public space, enterprise, or residential.
The main challenges in public space deployments are related to complexities in site
acquisitions, scaling of backhaul network, network/radio frequency planning, and
network management of large number of small cells. Other challenges are related to
managing spectrum choices based on location, subscriber services, and inter-
technology mobility.
The challenges in residential small cell deployments are similar to those observed in
public and enterprise deployments. The prevalent issues are related to 3G and LTE
spectrum choices based on residential location and subscriber services. From the
subscriber point of view it is better to have one physical unit rather than having
multiple units catering for multiple technologies.
Lack of new site locations, cost of new deployment and backhaul related aspects are
leading operators to see these 3G and LTE small cells under the same enclosure
sharing location, and backhaul. One possible integration model in an outdoor
deployment is a multi-mode small cell that combines 3GPP RATs with Wi-Fi
connectivity.
3G networks are going to co-exist with 4G networks for quite some time. As operators
embrace LTE rollouts, supporting current 3G deployments remains a need. Thus
overhead is apparent in CAPEX and OPEX and the added complexities involved in
procuring multiple sites or multiple site licenses, Radio frequency planning, and
multiple backhaul connections for integrating a carrier’s 3G and 4G small cell offerings.
Taking this into consideration, the overhead of building and maintaining separate
small cell networks for 3G and 4G should, on the face of it, be significantly reduced by
using integrated 3G and 4G small cell solutions. [1].
Some of the possible means for addressing the challenges involved in the multi-
technology small cell deployments are described below:
This section will present our view of alternative architectures for multi-technology
small cells. Those architectures will refer to the small cell as home eNode B (HeNB) for
4G or LTE small cells, and home node B (HNB) for 3G small cells.
3GPP already defines three possible variant options of network architecture for HeNB
Variant 1, 2, and 3 in [2], [3]. Those options differ in terms of whether a HeNB
gateway (HeNB-GW) is part of the network deployment or not, and if deployed
whether HeNB-GW aggregates both the control-plane (namely, C-Plane) and the user-
plane (namely, U-Plane), or it only aggregates the C-plane. More specifically, in
Variant 1 the HeNB-GW serves as a concentrator for the C-plane and also terminates
the U-plane towards the HeNB and serving gateway (S-GW). In Variant 2, the S1-U
interface of HeNB is terminated in S-GW and the S1-C interface is terminated directly
in mobility management entity (MME). HeNB may have connection to multiple
MME/serving gateway (S-GW). This represents the case of HeNB with S1-flex support.
In Variant 3, the HeNB-GW serves as a concentrator for the C-Plane only, and the S1-
U interface of HeNB is terminated in the S-GW.
Herein we discuss the multi technology 3G/LTE small cell network architecture options
considering all the three LTE architecture variants [4].
Note, the choice of which LTE architecture variant adopted by an operator may
be influenced by the requirements to support network sharing, as described in
[5].
The following four possible options for 3G/LTE small cell network architectures are
identified:
The following sections will describe Architectures A, B, C, and D in more detail. Any of
the three LTE architecture variants (1, 2, and 3) is applicable for these architectures.
In this option there is no functionality or interface that is shared between the different
technologies (LTE and 3G) as depicted in Figure 3-1. There is a need to identify and
separate the two types of traffic originating from LTE or 3G small cells. The combined
small cell will typically be allocated two separate IP addresses, one for use by the 3G
small cell (shown as HNB in the figure) and the other for use by the LTE small cell
(shown as HeNB in the figure)
In terms of mobility management, in all examples it is assumed that the core network
is able to support idle mode signalling reduction (ISR) incorporating the LAC/RAC
allocated to the 3G small cell and TAC allocated to the LTE small cell.
In this option a single management system for 3G and LTE is assumed, as illustrated
in Figure 3-2. This assumption fits within the current direction of the Broadband Forum
towards creating a unified way of managing multiple technologies [6]. A management
server provides the common FQDN for combined serving HMS/HeMS to HNB/HeNB for
the actual provisioning during this pre provisioning phase.
Combining the HNB management system (HMS) and HeNB management system
(HeMS) provides enhanced SON functionality and coordination between 3G and LTE
small cells, as the management system will have visibility into both the 3G and LTE
small cell networks. In particular, the combined management system is able to co-
ordinate neighbor cell lists (NCL) provisioned on both the 3G small cell and the LTE
small cell. This offers the management system a more complete view of the radio
access network (RAN) to make its decisions.
This architecture option could be applicable for the scenario where the element
management system (EMS) is perhaps already deployed for 3G and can additionally
support 4G requirements.
In this option a single security gateway (SeGW) is deployed for both 3G and LTE small
cells networks, as illustrated in Figure 3-3 (Architecture C-1) and Figure 3-4
(Architecture C-2). The common HMS/HeMS provisions the HNB and HeNB with a
common SeGW during the provisioning phase.
Figure 3-4 Architecture C-2 – Integrated security gateway (with a combined IPsec Tunnel for
luh and S1 interfaces)
In this architecture a single HNB/HeNB gateway is deployed for both 3G and LTE small
cells networks, as illustrated in Figure 3-5. The common HMS/HeMS provisions the
HNB and HeNB with a combined HNB-GW/HeNB-GW during the provisioning phase.
Architecture D should be more suitable for operators who already have a legacy 3G
deployment and plan to migrate towards LTE in order to provide the necessary cost
savings and hardware reuse flexibility.
3.5 Conclusion
For multi technology deployments, operators can either choose a single identified
architecture or combine the architectures described in this whitepaper based on their
legacy technology deployments and possible new technology requirements. For
example:
Overall multi-RAT RRM model should exploit the presence of the synergy between
multi-layer, multi-band, multi-vendor, multi-technology, and multi-service diversity in
a HetNet environment.
The ultimate multi-technology RRM objective is to satisfy the QoS requirements of the
individual radio bearers and to optimize the pool of available radio resources among a
variety of the multi-access technologies at the smallest possible cost for the network.
The trends in next-gen wireless network evolution indicate a desire to integrate a
variety of wireless access technologies for an ‘always best connected’ environment for
mobile users. This can be accomplished through efficient use of resources (for new
and existing users) through achieving the best QoS by minimizing call blocking,
dropped call rates and handover failures and maximizing overall network performance,
capacity and coverage at the minimum cost to the operators.
Multi-technology RRM functionalities can be split into the following three procedures,
as shown in Figure 4-2:
Examining the RRM functions shows that a high degree of common information is
shared by each RRM function across multi-RAT access points. In particular, a good
portion of measurements, reporting, and monitoring pool of resource procedures is
RAT agnostic. Note that, in multi-RAT RRM the decision enforcement procedure may
result in triggering other RRM functionalities. For example, handling the task of
uneven distribution (load balancing) of the traffic load over multiple inter-frequency
and inter-RAT cells can lead to handover and/or cell reselection decisions. Figure 4-3
illustrates the main functions and procedures of a multi-RAT RRM.
Although from the topology point of view, multi-RAT SON and multi-RAT RRM
implementation models might have a seemingly similar topologies in a HetNet, SON
functionality is more to do with the ‘self-’ configuration, optimization and maintenance
aspects of multi-cell/multi-RAT networks.
One possible architecture for efficient multi-RAT RRM is a two-tier RRM architecture
similar to that shown in Figure 4-3.
Local RRM: this RRM entity physically resides within each access point (AP) or User
equipment (UE). It manages, allocates, and de-allocates RAT-specific radio resources
for the given radio technology network [10].
Multi-RAT RRM: this common RAT or controlling RRM entity has the knowledge of the
overall radio resource pool, and is responsible for efficient management of multiple
local RRM entities as well as exchanging information reporting with other multi-RAT
RRMs
Based on the degree of interaction between local RRM and multi-RAM entities, the
following functions can be performed by either local RRM or multi-RAT RRM entities:
In general, the multi-RAT RRM and the local RRM will work in a hierarchical manner,
where the resources managed by the local RRM will depend on the assignments
performed by the multi-RAT RRM, which will depend on feedback information coming
from the local RRM and external policies.
The level of granularity of local RRM, which typically contains the scheduler, will be
usually smaller than the multi-RAT RRM. For example, LTE and LTE-A can define
scheduling decisions in the order of 1ms (TTI) that is more appropriate to be
implemented in local RRMs. Multi-RAT RRM is typically operational in seconds or
minutes granularity.
The use of a multi-RAT RRM as a central entity could only be required and used when
the local RRM entities are not able to further fulfil the network and user requirements.
Most of the interactions between local RRMs and multi-RAT RRMs are on the low or
intermediate time-scale levels. This allows the multi-RAT RRM entity to perform
functions such as RAT selection, vertical handover (inter-RAT), admission control, and
congestion control functions. A higher degree of interaction between local RRM and
multi-RAT RRM entities can achieve more efficient resource management. However, it
requires more frequent interactions between these entities, which may lead to a
higher amount of signalling overhead.
The interaction between the local RRM and multi-RAT RRM is not directly related to the
coupling topologies between radio access networks. Nevertheless, for frequent local
RRM and multi-RAT RRM interactions, tight coupling topologies are required in order to
reduce the delays in communications between levels of hierarchies. Various degrees of
coupling between multi-RAT RRM and local RRM entities in terms of short/long-term
time scales ranging from less than 1ms (e.g., inner loop power control in WCDMA) to
thousands of frames (e.g., admission control, Handover or outer loop power control)
can occur. In general, the more frequent interactions between multi-RAT RRM and
local RRM, the tighter the coupling architectures are required, as illustrated in Figure
4-4. The lowest degree of interaction occurs when the multi-RAT RRM is only
responsible for dictating operator’s policy enforcement as various configurations for
lower local RRM functions.
From the implementation point of view and coupling architectures, the simplest
solution for heterogeneous networks integration is the so-called ‘loose coupling’
architecture. In this implementation, different networks are connected together
through gateways, while still maintaining their independence. In general, a number of
possible configurations can be considered for multi-RAT RRM implementations:
• Centralized RRM
• Distributed RRM
• Localized RRM
• Hybrid RRM
• Some RRM resides in the operation and maintenance (OAM) servers and
some on network management (that is, in eNB).
In general, distributed or hybrid approaches are favoured since they enable lower
delay, less signalling, and lower cost, even though they risk losing some performance
gains compared to their centralized counterparts.
The Multi-RAT RRM topologies can be, possibly, split into the following models based
on the implementation constraints and the use cases as shown in Error! Reference
source not found.:
Use
Multi-RAT RRM
Case Description Implications
implementation models
#
• Integrated single • Localized • No new interface/standard is
access point multi-RRM required within the APs
• Cell or AP centric • All multi- • Decision processes are made
technology locally
stacks are in • New interfaces may be required
1
the same for information reposting across
access point APs (inter-RAT)
• Dynamic RRM handling, requiring
frequent signal exchanges
Table 4-1 Various Multi-RAT RRM models and associated use cases
In an integrated multi-RAT RRM single access point, the execution of the multi-RRM
functions can be performed locally between various RATs rather than through multiple
APs or CNs. In this topology, no additional delays will be incurred. Since the RRM
decision processes between RATs are co-located within the same physical entities, the
support functions do not need to be standardized and no new open interfaces need to
be defined. The entire RRM process is a cell or a single access point (AP) centric
(intra-AP communication). See Figure 4-5 for an integrated multi-RAT RRM topology.
Figure 4-6 shows various possible scenarios for an integrated multi-RAT RRM across
inter-RAT APs. In scenario (a), the reporting information between intra-RAT can be
communicated through standard interfaces such as X2, Iur. In the case of inter-RAT
RRMs such as UTRAN and GERAN, the communication between BSC and radio network
controller (RNC) NC can be performed through the MSC of the core network. One the
other hand, in the case (b), where multi-RRM entities are not included in some of the
access points, either new interfaces need to be defined or the communication between
the multi-RRMs should be performed through the core network, using the existing
interfaces that are defined in the 3GPP standards (e.g., S1,Iu, Iuh, Iur-g, Gb). Finally,
in the case (c), a combination of an integrated inter-RAT RRM in a single AP and
another inter-RAT in another AP is shown. Note that the dashed-line interface
connections may or may not exist. In case that the dashed interfaces are missing the
communications between APs are performed through the core network (or using a
cloud RAN).
In a server-based multi-RAT RRM topology, a server may reside in the core network.
The server-based multi-RAT RRM is centralized and common to multiple technologies
This topology allows a greater part of the RRM decisions to be assigned to the UE
including making the RAT selection decisions. User-centric RRM typically assigns
resources to users first and then to radio ports. This approach requires more
computation and power consumption from the UE, in addition to information from the
RAN and core network. Figure 4-8 illustrates a UE-centric multi-RAT RRM topology.
The advantage of having a hierarchical RRM is that it allows the lower-level entities in
the hierarchy to perform and communicate the RRM decisions faster and with less
overhead than in a scheme that only depends on a central RRM entity. However, in
the case of low network load or low number of local RRM entities, the use of multi-RAT
RRM can be avoided if the local RRMs are available and capable of managing the
4.7 Conclusion
Multi-RAT RRM can bring significant benefits in the heterogeneous network including
load balancing, reduced interference distribution, minimized unnecessary handovers,
Depending on the topology and the coupling interaction between local RRM and multi-
RAT RRM, the amount of signaling overhead may increase. To this point, trade-offs
between multi-RAT RRM architecture, performance, strategies and signaling overhead
are required to find the optimum solution.
Although the Multi-RAT RRM algorithms do not need to be standardized, the issue of
the interoperability between multi-vendors and multi-RAT RANs could be addressed
through a set of defined (open) interfaces for a full deployment of always best connect
future networks.
Small cells in the form of femtocells, picocells, or microcells are strategically placed to
offer data offload from the existing macro network and provide coverage and capacity
improvement.
Multi-technology small cells could enable traffic offload with effective resource
handling, and load balance across multiple accesses.
Multi-technology small cells offer additional use cases for traffic offload and traffic
steering: case in point, use of 3G for voice calls, prioritizing of traffic for premium
services/users.
Local IP access (namely, LIPA) enables an UE connected via a small cell to access
other IP-capable devices in the same residential/enterprise IP network without the
user plane traversing the mobile operator’s network except the small cell subsystem.
The local IP Access is achieved using a local GW (L-GW) collocated with the small cell.
The enterprise small cell network architecture, that includes an optional enterprise
small cell gateway (E-SCG) for enabling offload of enterprise traffic directly onto the
Intranet, is described in details in [12]. The E-SCG incorporates the 3GPP defined
local GW (L-GW) functionality.
For multi-technology small cells, a single instance of E-SCG can offer LIPA functionality
for both 3G and 4G technologies. Figure 5-1 shows an example LIPA for multi-mode
small cells.
Network offload gateways provide a managed traffic offload option for operators that
use the principle of offloading heaviest traffic before it reaches the core network.
For Multi-technology small cells, a single instance of network offload gateway can offer
SIPTO functionality for both 3G and 4G technologies. This mechanism can be offered
as an edge functionality to deliver capability as value-add to gateways deployed at the
edge of the network. As we review the multi-mode network architecture, this offload
function can be at the common small cell GW for 3G/LTE enabling traffic offload.
Note: For more insight into L-GW and SIPTO capabilities refer to Error!
Reference source not found..
• Operator benefits:
This paper explores possible technology synergies that could serve as integrated
solutions for future multi-technology small cell deployments. This includes feasible
multi-technology small cell architectures and their pros and cons in terms of
optimization of product integration and signalling, system robustness, security,
flexibility and scalability for future upgrade, and networks saving in CAPEX and OPEX.
Trade-offs between multi-RAT RRM architecture, performance, strategies and
signalling overhead are required to find the optimum multi-RAT RRM solution for
future multi-technology small cell networks. The integrated multi-RAT radio resource
management solution, highlighted in this paper, could be proposed for future multi-
technology small cell deployments. Nonetheless, any foreseen benefits to the use of
this solution in terms of load balancing, interference mitigation, or possible reduction
in signalling overhead, etc., shall be evaluated based on individual operator’s
deployment use cases and offered services. In terms of traffic offload, the paper
proposes possible multi-technology small cells deployments that enable traffic offload
with effective resource handling, optimal mobility and load balancing capability across
multiple access networks.