Computer Science

Omar Amoretti

Unleashing VoLTE capabilities.Assessing

the Migration from CS Voice to IMS-based
Voice over LTE (VoLTE)

Master's Thesis
Bibliographic information published by the German National Library:

The German National Library lists this publication in the National Bibliography;
detailed bibliographic data are available on the Internet at http://dnb.dnb.de .

This book is copyright material and must not be copied, reproduced, transferred,
distributed, leased, licensed or publicly performed or used in any way except as
specifically permitted in writing by the publishers, as allowed under the terms and
conditions under which it was purchased or as strictly permitted by applicable
copyright law. Any unauthorized distribution or use of this text may be a direct
infringement of the author s and publisher s rights and those responsible may be
liable in law accordingly.

Imprint:
Copyright © 2016 GRIN Verlag, Open Publishing GmbH
ISBN: 9783668219854

This book at GRIN:

http://www.grin.com/en/e-book/322492/unleashing-volte-capabilities-assessing-the-
migration-from-cs-voice-to
Omar Amoretti

Unleashing VoLTE capabilities. Assessing the Migration

from CS Voice to IMS-based Voice over LTE (VoLTE)

GRIN Publishing
GRIN - Your knowledge has value

Since its foundation in 1998, GRIN has specialized in publishing academic texts by
students, college teachers and other academics as e-book and printed book. The
website www.grin.com is an ideal platform for presenting term papers, final papers,
scientific essays, dissertations and specialist books.

Visit us on the internet:

http://www.grin.com/

http://www.facebook.com/grincom

http://www.twitter.com/grin_com
Master thesis

Unleashing VoLTE capabilities

Assessing the Migration from CS Voice to IMS-
based Voice over LTE (VoLTE)

By: Mag. Omar AMORETTI, MSc

Vienna, 2016
Abstract
Operator-provided voice services will gradually migrate from today’s circuit-switched (CS)
voice networks to packet-switched IP networks, using Voice over LTE (VoLTE) as the
foundation to provide telecom-grade telephony services. This paper presents the first
implications of introducing IMS-based VoLTE in Germany, Austria and Switzerland (DACH)
from a mobile operator’s perspective. Further VoLTE performance aspects such as
parameter optimization and HD voice are discussed, thus serving as a basis to then analyze
the DACH mobile network test results for 2014 in terms of telephony and data performance.
Overall results for a DACH VoLTE trial conducted in the third quarter 2014 complement the
analysis.

Keywords: VoLTE, IMS-based VoLTE, HD voice with VoLTE

Table of Contents

1  Introduction .............................................................................................................. 4 

2  Migrating to IMS-based VoLTE: Initial considerations ............................................. 7 

2.1  Benefits of IMS-based VoLTE ................................................................................ 10 
2.2  Providing voice over LTE ....................................................................................... 12 
2.2.1  Circuit-switched Fallback (CSFB) .......................................................................... 14 
2.2.2  Voice over LTE (VoLTE) ........................................................................................ 16 
2.3  VoLTE Logical Architecture ................................................................................... 20 
2.3.1  4G LTE network architecture ................................................................................. 20 
2.3.2  VoLTE Functional Node Description ...................................................................... 21 
2.4  Assessing the Voice Quality Performance ............................................................. 25 
2.4.1  Leveraging the Operators’ Assets .......................................................................... 26 
2.4.2  Development of audio testing ................................................................................ 28 
2.4.3  PESQ/POLQA methodology .................................................................................. 30 

3  VoLTE Parameter Optimization ............................................................................. 32 

3.1  Robust Header Compression (RoHC) ................................................................... 33 
3.2  Transmission Time Interval (TTI) Bundling ............................................................ 34 
3.3  Discontinuous Reception (DRX) ............................................................................ 35 
3.4  Dedicated Bearers ................................................................................................. 36 
3.5  Semi-Persistent Scheduling (SPS) ........................................................................ 37 

4  VoLTE Testing ....................................................................................................... 40 

4.1  Primary Implications ............................................................................................... 40 
4.1.1  Voice Quality of VoLTE versus OTT Voice Services ............................................. 41 
4.1.2  Mobile Network Test: Testing Framework ............................................................. 43 
4.2  Mobile Network Test 2014: Connect Magazine ..................................................... 46 
4.2.1  Organizational aspects .......................................................................................... 46 
4.2.2  Test methodology .................................................................................................. 47 
4.2.3  Germany: Telephony Test Results ........................................................................ 49 
4.2.4  Germany: Data Test Results .................................................................................. 51 
4.2.5  Austria: Telephony Test Results ............................................................................ 53 
4.2.6  Austria: Data Test Results ..................................................................................... 54 

2
 4.2.7  Switzerland: Telephony Test Results ..................................................................... 56 
4.2.8  Switzerland: Data Test Results .............................................................................. 57 
4.2.9  Mobile Network Test: Overall Results DACH Region ............................................ 59 
4.2.10  Austria: VoLTE Friendly User Tests ....................................................................... 60 
4.3  Mobile Network Test 2014: CHIP Magazine .......................................................... 61 
4.3.1  Test methodology .................................................................................................. 62 
4.3.2  Germany: Telephony Test Results ........................................................................ 62 
4.3.3  Germany: Data Test Results .................................................................................. 64 

5  Conclusion ............................................................................................................. 66 

Bibliography ........................................................................................................................... 68 

List of Figures ........................................................................................................................ 71 

List of Tables ......................................................................................................................... 72 

List of Abbreviations .............................................................................................................. 73 

3
1 Introduction
The increasing gap among capacity and demand poses an urgent call for novel network
technologies to allow mobile operators to improve performance on a cost-effective basis. In
fact, Voice over Long Term Evolution (VoLTE) is a key component for an innovative set of
services determined for all-IP networks: the goal relies on making these novel applications
as available as voice and messaging are nowadays, while also offering a flexible
interaction with Internet applications.

Indeed, LTE focus on a (rather flat) all-IP access technology aiming at delivering a
bandwidth-efficient approach for carrying several types of subscriber traffic at the same
time. In other words, the capability of transporting Voice over IP (VoIP) services along with
the provision of high-rate data throughputs, characterizes one of the critical drivers for the
development to LTE. As of December 2014, six countries have deployed VoLTE1 services
and still 56 others are currently preparing its implementation [30].

VoLTE denotes a GSMA standards’ profile for the provision of applications offered
nowadays through Circuit Switched (CS) networks over the Packet Switched (PS) network
of LTE. For many operators, VoLTE embodies a future-proof approach to rich media
services, thus leaving no other reasonable alternative than to opt for its swift
implementation. In this sense, the VoLTE deployment has underlined the importance of
realizing the IP Multimedia Subsystem (IMS) and its related Session Initiation Protocol
(SIP) within a wireless setting. Certainly, IMS and SIP are crucial in enabling VoIP
services like VoLTE in a LTE environment.

1
Additionally, 125 mobile network operators (MNO) in 77 countries have commercially deployed HD voice services by the
end of the first quarter 2015 [30]

4
For instance, IMS delivers the gateway functionality and interconnection supporting the
communication between VoIP and non-VoIP devices2. For its part, SIP determines the
required signaling for authentication, call establishment and the delivery of complementary
services3. Undeniably, to deliver ‘telecom-grade’ voice services (at least better than the
legacy CS ones) is of outstanding relevance to VoLTE. In this regard, the VoIP
implementation in a LTE setting clearly requires both technologies (SIP/IMS) along with
LTE radio access network (RAN) features, which altogether enable VoLTE to outperform
over-the-top (OTT) voice services [28].

Generally speaking, these OTT applications involve the provision of media (audio/video) to
a subscriber’s handset and rely on the specific Internet service provider (ISP) only as a
carrier of bits and bytes. Definitely, OTT denotes not only streaming services4, it also
comprises VoIP speech applications.

Within this context, VoLTE5 is predestined to turn into the key technology for IP-based
communications. What is more, VoLTE enables even better high definition (HD) voice
quality, bringing forward the mobile network operator’s (MNO) competitive capabilities with
regard to OTT VoIP providers. In the same manner, the MNO’s strategy for voice as well
as spectrum availability will ultimately shape the VoLTE’s deployment. As operators define
their implementation plans, the Circuit Switched Fallback (CSFB) might represent a first
evolutionary approach, followed by the Single Radio Voice Call Continuity (SRVCC) as
soon as VoLTE is launched6 [35]. Alternatively, other MNOs would await eagerly for
ubiquitous LTE coverage before their VoLTE service offering. In fact, the MNO’s options
will be affected by specific business goals, its technology architecture as well as strategic
competences.

Rich Communications Services (RCS) will (analogous to VoLTE) benefit from IMS control
capabilities, thus embracing an enriched multimedia portfolio. For instance, starting with
the first VoLTE launch in 2012 in South Korea, a comprehensive variety of VoLTE-capable
devices has been developed, leading many operators to launch it during 2014 (Asia and
North America [30].

2
non-wireless devices also included
3
e.g. call retention or three-way calling
4
Hulu, WhatsApp, Netflix, etc.
5
together with a complementary conversational video service defined by GSMA
6
prior to ubiquitous LTE coverage

5
Furthermore, the numerical data throughout presented in this document uses a comma ","
as decimal mark. Moreover, this paper is divided in four parts, the first of which describes
some key implications of the migration to IMS-based VoLTE (chapter two). Undeniably,
deploying VoLTE requires a number of optimization steps to gain the full benefit of the
technology’s potential. This parameter optimization is the focus of chapter three. The fourth
chapter presents a high level analysis of two mobile network tests published in 2014 that
builds the basis for further findings regarding telephony and data (access) performance.
Particularly, this sheds light not only on the voice test methodology, but also on the
importance of VoLTE Friendly User Tests to better align the operator’s voice capabilities
(over LTE) with the demanding user’s expectations. Finally, some conclusions are drawn
based on the relevant aspects discussed, outlining the upcoming challenges operators will
cope with in the near term.

6
2 Migrating to IMS-based VoLTE: Initial
considerations

The need for bandwidth derived from devices and subscribers has been constantly
increasing for many years. Actually, the data volume transferred by mobile networks is
doubling approximately every year and the quantity of connected machine-to-machine
(M2M) devices is estimated to exceed 50 billion by 2020 [25]. Nevertheless, the compound
annual growth rate (CAGR) is expected to decline in the years to come, from 4% per
annum for 2008 – 2014 to a moderate 3,1% p.a. till 2020 [30].

In this sense, to take advantage of mobile-broadband opportunities (while at the same time
increasing benefits for business and end-users) remains the focus of every operator’s
activities. LTE networks, for their part, are able to carry mobile broadband with huge data
capacity and a minor latency level. Though, since there is no circuit-switched voice domain
in LTE (Fig. 1), a universally interoperable IP-based voice and video calling solution for
LTE is being implemented within the telecom industry: VoLTE [29] and ViLTE7, which
further facilitate the evolution of innovative communication services.

Figure 1: Circuit and packet domains. LTE within the Evolved Packet System (EPS) [1]

Furthermore, over-the-top (OTT) communication solutions like Skype (and most recently
WhatsApp) have influenced the way users assess a particular service based on VoIP.
According to an estimation [26] [30], the business loss derived from users using OTT voice
services will amount to USD 386 billion for the period 2012 - 2018. Nonetheless, a
completely satisfactory user experience cannot be offered by OTT solutions due to missing
QoS measures or the lack of handover mechanisms to the circuit-switched (CS) network.

7
A conversational video application founded on the IP Multimedia Subsystem (IMS) and defined by GSMA IR.94 [29]

7
In addition, there is no guaranteed emergency call support or extensive interoperability of
services among diverse OTT services. Thus, the readiness of subscribers to use a service
that does not offer security, quality, flexibility or even mobile-broadband coverage clearly
influence the adoption of OTT services in a negative manner.

Furthermore, VoLTE is expected to be implemented together with CS voice, e.g. parallel to

the ongoing evolution towards IMS-based VoLTE: its timing becomes vital for operators to
prevent losing out to OTT providers and make the most out of VoLTE capabilities.
Operators, therefore, apply the following methods:

 Spectrum refarming8 allows to considerably limit the implementation expenses and

supports twice as many calls within the same spectrum (e.g. spectral efficiency).
 Deployment expenditures can be substantially decreased in contrast to spectrum
acquisition
 Network and planning optimization poses a cost restriction and facilitates future-
proof network modernization

The past two years have seen a wave of LTE network deployments around the world [30].
Interestingly, most mobile network operators (MNO) still use a CS system because of their
reliability in carrying voice services and also because of user preferences (many still use
CS terminals). For instance, many MNOs are currently evaluating a suitable path to ‘align’
their networks with the Internet by migrating the entire services to a (LTE IP) packet-
switched environment. For this purpose, the IP Multimedia Subsystem (IMS, [12]), a
system standardized by the 3rd Generation Partnership Project (3GPP), has proven to be
the commonly chosen solution.

8
It can be considered as a process involving any basic change in the terms of frequency usage for a specific part of the radio
spectrum

8
The clear advantage for IMS is its cost-effectiveness when shifting core telecom services
(voice, video and messaging) to an IP environment. Additionally, IMS is more network-
efficient than current systems and generates supplementary prospects for innovation in the
medium term. Particularly, the basic IMS capabilities9 can be reused by further services,
thus rendering IMS a trend-setting and comprehensive solution in regard to a broader IP-
communication approach. Besides, IMS is ‘access-agnostic’, meaning the support for
seamless handovers and its independence of the connection method.

Regulatory measures
In general terms, the European Union follows a consumer-oriented as well as
interventionist style. In fact, the drop of termination rates, e.g. the fees that one MNO
charges another one for terminating calls on its network, has brought a significant
weakening of the MNO’s interconnection earnings. In this sense, the European
Commission adopted by the end of 2013 a Recommendation [23] to the national regulatory
authorities regarding the non-discriminative practices on costing methodologies in order to
encourage competition and boost the investment in broadband infrastructure. Still, the
European Commission proposed in the same year an EU Regulation targeted at the further
development of the single market for electronic communications.

Furthermore, the EU strives to eradicate the roaming fees10 for consumers within Europe.
The ‘Digital Single Market Proposal’, presented by the European Commission two years
ago, would involve a decline in roaming fees for messages, data and calls to the particular
domestic termination rates level (‘roam like at home’) [23]. Irrespective of the proposal’s
result, this sort of harmonization is projected to happen in the medium term, causing
returns to further deteriorate.

Moreover, competition authorities impose rigorous controls on telecom players with the aim
of guaranteeing the (MNO) compliance with industry specific regulations. It is, therefore,
indispensable that European network operators mitigate the negative profit effects owing to
price-oriented competition and regulation by concentrating on profitable innovations and
(market) sectors. Undoubtedly, finding the appropriate equilibrium among (infrastructure)
capital expenditures and cost-efficiency is a crucial element to gain and maintain
momentum.

9
authentication, authorization, charging and routing
10
a plan currently under discussion

9
Undeniable, a well thought-out (European Union’s) regulatory method can fully benefit of
the socio-economic impact derived from the mobile business by facilitating innovation in
mobile connectivity and applications. So, the prospect of a ‘connected’ Europe is better
accomplished by promoting investments as well as end-user confidence in novel mobile
technologies.

2.1 Benefits of IMS-based VoLTE

In the medium term, VoLTE operators have a sustainable competitive positioning: they can
deliver superior HD voice, implement video/messaging as well as converge with the web
via Web Real Time Communications (WebRTC). Moreover, (VoLTE) operators can work in
partnership with application providers and supply the best user experience on a low cost
basis, because they can translate application developers’ solutions onto innovative
communication services. Also, a quick customization of features for strategic activities (as
the mobile healthcare industry could be) becomes feasible and effective.

COMPETITIVE  •Fully monetize the 4G LTE

investment
VOICE, VIDEO 
•Reduce the communication
AND  price erosion and increase
MESSAGING broadband profits

•Experiment and invent new

services
INNOVATION
•Service agility enables a
continued differentiation

•Control costs and improve

service delivery
CONVERGED 
•Embrace the subscriber’s life
CLOUD across the web, fixed and
mobile world

Figure 2: VoLTE’s strategic value for operators. Adapted from [14]

10
Needless to say, VoLTE plays an important role in empowering all-IP communications in
the 4G LTE network. As a result, VoLTE will enable operators to:

 Generate attractive communication services by merging mobile voice with video,

the web and social networking
 Provide faster call setup times
 Improve customer experience by delivering data and better HD voice
simultaneously (which also contributes to offload legacy infrastructure). In addition,
the current ‘fragmented’ communications landscape that relies on various (rich)
media approaches becomes more harmonized
 Migrate from dual radio CDMA (only in the USA) and LTE devices to LTE-only
devices

Furthermore, the user value proposition for IP-communications11 relies indeed in the
MNO’s traditional assets: while each service (RCS, VoLTE) delivers some real benefits
compared to other approaches, the intrinsically user value proposition of IP-
communications is not based on pioneering service enrichments, but rather on a superior
user experience centered on reliability and interconnection as well (refer to Fig. 3).

Figure 3: Call setup time measurements in comparison [36]

11
which is comprised of VoLTE and IP-messaging or Rich Communication Services (RCS)

11
For sure, VoLTE is not the only method by which all operators will initially launch voice
service in the short run. Yet, it embodies the ideal approach if competitive advantages and
their related (business) risks are considered.

2.2 Providing voice over LTE

Despite the overwhelming enthusiasm on just exploiting LTE for data, several MNOs
consider the provision of voice services as LTE is being implemented into the network. In
an initial stage, some approaches involve the deployment of LTE to complement the
already existent High Speed Packet Access HSPA/HSPA+ and EDGE coverage (within
data-intensive areas for example). Additionally, the phased deployment of LTE throughout
the whole network forces operators to safeguard the continuity of voice services (with slight
service interruptions) as long as subscribers roam among 2G/3G and LTE networks. In the
end, since that there is no other well-specified, wide accepted or well-supported solution,
operators will have no other choice than adopting VoLTE.

For its part, 3GPP provides two standards for voice service provision in 4G LTE [3] [5]:

 VoLTE: Here, the LTE network (and IMS) care for the delivery of voice calls
 Circuit Switched Fallback (CSFB): Given an established LTE connection, the User
Equipment (UE) has to fall back to 2G/3G the moment a call origination/termination
occurs

Particularly, the CS Fallback (CSFB) and the dual-radio methods on which CDMA
operators rely (in some cases known as Simultaneous Voice-and-LTE, SVLTE) are
convenient to the extent that current telephony services are reused. However, these are
not indeed LTE solutions and, as a result, are not subject to suitable development.
Likewise, OTT methods are attractive since they run over IP and support, on the one hand,
richer experiences. On the other hand, important features like interoperability and 2G/3G
integration/handover are
(with the OTT approach) definitely at stake (see Table 1).

12
 Subscriber’s Service VoLTE CSFB App
Providers
Standards Global interoperability, including regulatory Yes Yes No
End to End QoS Yes Yes No
Roam with local voice, not home-routed data Yes Yes No
Multimedia 4G LTE data simultaneous with voice Yes No Yes
All-IP network enables video-communications, etc. Yes No Yes
Foundation for services innovation, WebRTC , etc. Yes No Yes
Voice Minimal voice call setup delay Yes No Yes
Evolved voice: HD, new features, WebRTC , etc. Yes No Yes
Graceful continuity to 2G/3G circuit voice Yes Yes No
Table 1: VoLTE compared to other approaches for voice in LTE devices. Adapted from [16]

The development path towards VoLTE, as shown in Fig. 412, may follow different
approaches: without a doubt, some operators will opt for a blend of CS voice and LTE data
(on LTE devices). Sooner or later, more profitable prospects derived from enriched voice
services for LTE smartphones will persuade most MNOs to fully adopt IMS-based VoLTE.
By delivering voice and data over the same network at the same time, a superior customer
experience can be accomplished.

It should be noted that the features influencing the migration alternatives (e.g. rollout,
coverage as well as the implementation speed) are not only determined by the present
mobile technology or market conditions, but ultimately by the available spectrum in a
particular site.

12
VoLGA is not included in the graph since the standards have yet been accepted by 3GPP. Hence, it is outdated

13
 Figure 4: Voice Strategies: Providing voice over LTE [35]

2.2.1 Circuit-switched Fallback (CSFB)

CSFB embodies an initial stage in empowering LTE devices with the size, cost-
effectiveness as well as power consumption benefits of single-radio approaches (in terms
of 2G/3G voice combined with LTE data). In fact, CSFB reuses the legacy networks in
order to deliver voice services for LTE. For this purpose, the handset13 is required to fall
back to the legacy network prior to the call setup origination/termination. In fact, CSFB
provides feature transparency and a comprehensive service while facilitating operators to
exploit their 2G/3G networks for the CS service provision.

13
usually attached to the LTE network for data services

14
At first glance, some benefits and drawbacks derived from CSFB are outlined in Table 2
Benefits Drawbacks
CSFB prolongs the lifecycle of 2G/3G networks CSFB is fairly signaling-intensive and the time
by allowing them to support voice services for to complete the fallback (when both sides
the LTE network perform it) usually takes around 6 seconds
(Fig. 3).
Many 2G/3G elements like CS service platforms, An extra call setup delay (due to the fallback)
operations support systems (OSS), MSC, etc. occurs when the handset performs
are reused, thus guaranteeing a quick and measurements to locate an appropriate
reliable voice service rollout for LTE. GSM/UMTS cell to use. Still, a location update
is necessary before the origination of (or the
reply to) a call.
For CSFB, no additional network components CSFB provides simultaneous voice and data
are required. Also, upgrades to current network on handovers to UMTS and GSM with Dual
nodes are rather minor versus other options Transfer Mode (DTM). However, this is not the
case when handing down to a 2G network
lacking DTM: here, the PS session gets merely
canceled.
CSFB modifications to the MSC are indeed Increased OPEX and risk: the economic and
simple since the SG interface was intentionally technical challenges that rise with the
founded on the Gs interface (employed among operation and maintenance of 4G LTE and
the MSC and Serving GPRS Support Node). legacy 2G/3G networks in parallel should not
be underestimated.

Table 2: Assessing the CSFB approach

Nevertheless, operators who temporarily choose CSFB should critically reflect on the
invested resources due to:

 Communication experience for end-users might be at stake: during voice calls

CSFB subscribers would be downgraded from 4G LTE data service to 3G HSPA+
or 2G, losing data services completely.
 Boundaries are imposed on innovation: New services, which work on all-IP
networks and are based on high quality video codecs, cannot be realized (such as
video calling)
 In the medium term, CSFB has to be supported for devices lacking VoLTE
capabilities and for inbound roaming subscribers

15
Another aspect to consider is the supply of devices via non-operator sales channels
(‘vanilla’14 handsets), which may negatively affect the supported voice features. Even if
carrier-controlled handsets will predominate, complex configuration steps would be
necessary for those not provided by the MNOs.

2.2.2 Voice over LTE (VoLTE)

This approach relies on the IMS call control as stated by 3GPP TS 23.228 [12] for LTE
voice application carriage. In fact, IMS delivers value-added innovative multimedia services
by facilitating media additions and removals at whatever time throughout a call. Likewise, it
is estimated that VoLTE will be extensively implemented in view of the wide-ranging LTE
coverage, regardless of the visited network.

Furthermore, end-to-end IP real-time voice and rich media services are allowed by VoLTE
executed with HSPA+ and LTE. Indeed, the benefit of this configuration is given by the
seamless mobility among LTE and UMTS by means of packet-switched (PS)
handovers/handbacks. For instance, VoLTE users would enjoy steady voice services
independent of their attachment to a 2G/3G or LTE network. This is guaranteed by the IMS
Centralized Services (ICS), in the case of 2G/3G CS service continuity, by linking the IMS
call control to the GSM/UMTS access network. Despite these advantages, most operators
do not follow the HSPA approach since it is less efficient for voice and QoS enforcement
(e.g. delay).

Single Radio Voice Call Continuity (SRVCC)

Normally, a conversation would be dropped once the LTE signal get lost and no handover
(towards 2G/3G) takes place. SRVCC and enhanced SRVCC15 [2], standardized by the
3GPP, characterize in this context the VoLTE technology responsible for this handover as
a way to guarantee the call continuity as soon as the end-user finds himself outside the
LTE coverage. In other words, SRVCC stands for the call continuity among IMS (over
PS/CS access) for conversations attached to IMS given that the handset is only capable of
communicating on one of those access networks [17]. What is more, SRVCC performs
network-controlled handovers, thus removing the necessity for the UE to connect to two
access networks at the same time.

14
e.g. without any (operator-side) customization
15
This applies for international roaming. SRVCC is defined by the 3GPP standards TS 24.237 [7], TS 23.237 [5] and TS
23.216 [4]

16
From the abovementioned, the mobile device is not able to attach to more than one radio
access networks simultaneously: this implies a cutting off from the LTE network and a
reconnection to 2G/3G. Should this arise during a chat, the SRVCC determines the actions
to be undertaken (e.g. SRVCC is not valid outside of a conversation). Another relevant
factor is given by the IMS call control in the home network of the end-user, which implies
that the (VoIP or CS-based) conversation is camped in the IMS domain.

Based on Fig. 5, the general SRVCC mechanism comprises an active VoLTE conversation
represented by step 1, followed by a SRVCC handover (after abandoning LTE coverage):
the Evolved Packet Core (EPC) ‘talks’ to the Mobile Switching Center (MSC), which
initiates the transfer from LTE towards 2G/3G (step 2). Additionally, the MSC prompts the
IMS to run a session handover, thus launching a new access route with the MSC. In the
last third step, the IMS-controlled conversation is reinitiated over the CS core.

Figure 5: Common SRVCC mechanism [17]

Moreover, the SRVCC procedure implies not only a handset having a SRVCC client, but
also requires the involvement of the Mobility Management Entity (MME), LTE’s radio part,
Service Centralization and Continuity Application Server (SCC-AS), the Home Subscriber
Server (HSS) as well as SRVCC-upgraded MSCs). These components are further
discussed in section 2.3.

IMS Centralized Services (ICS)

For the past years, IMS mobile telephony has been implemented as a response to the
growing subscriber demands and the advent of novel 4G technologies. This is because
IMS facilitates fast developing and comprehensive services above and beyond speech

17
(video for example). Since LTE16 is not yet available throughout the world, roaming for
mobile IMS rollouts still embodies a challenge. Fortunately, 3GPP and GSMA have
promptly react by outlining methods (e.g. IMS Centralized Services [15]) for the provision
of worldwide accessibility of IMS-controlled speech applications17. Nevertheless, no
general method has been defined, which implies an explicit mechanism for a particular
network type.

The term service consistency (SC), in the ICS context, points out the access reliability of
speech applications through 2G/3G and LTE networks. Particularly, SC comprises a wide
range of calls, even those terminated/originated from an IMS subscriber for the duration of
2G/3G network coverage. For the sake of clarification, Fig. 6 illustrates the difference
between SC (the field of ICS) and service continuity.

Figure 6: Service continuity compared to service consistency [15]

Furthermore, IMS-capable networks18 (or those roaming to a capable one) enable IMS
operators to offer separated (value-added) services to their subscribers. This may be
defined as the home control pattern. However, many networks neither provide IMS VoIP
(due to the absence of QoS controls) nor are IMS-capable, leaving one alternative for the
provision of voice: the application of 2G/3G circuit-mode MSCs. In this case, a defined
home-network pattern applies.

16
employed to transport IMS
17
even in CS networks
18
An application server (AS) located in the home network performs service logic. Exceptions of the service logic apply for
example for emergency services

18
To address the kind of incompatibility between the two settings (patterns), two options
appear reasonable: the first one involves the selection of voice services that can be reliable
provided via 2G/3G networks, while still offering superior speech applications through LTE.
On the other hand, a set of features offering IMS home control to subscribers assisted by
MSCs (3GPP ICS, [15]) may apply as the second alternative. Interestingly, most MNOs are
opting for the first alternative as no reasonable method has been identified for the second
one.

Table 3 examines further implications related to VoLTE

Benefits Drawbacks
VoLTE empowers MNO to launch novel profit The adoption of ICS and SRVCC (in legacy
generating voice/data services (GSMA Rich networks) would call for the arrangement of a
Communication Suite for example). Service Media Gateway Control Function (MGCF) to
providers, from their part, are more tempted to support the Sv interface19. Similarly,
provide converged fixed/mobile services on their improvements to each MSC contiguous to
wireless as well as wireline networks. LTE/IMS networks might become
indispensable
VoLTE make the most of the LTE/IMS network, The LTE network demands a considerable
bringing the operational cost reductions of investment owing to the introduction of the IMS
relying on a flat IP network. core and the IP-SM-GW20.

Simultaneous voice/data handovers (based on Due to the complexity of the SRVCC signaling,
VoLTE realized with LTE/UMTS PS access) a break in the voice stream may eventually be
become feasible due to their management over perceived by LTE subscribers (the case being
a single PS domain. a call ‘handed-down’ to the 2G/3G network).
The enhanced SRVCC (eSRVCC, 3GPP
Release 10) addresses this issue
Subscribers would experience equal voice
services on the LTE or 2G/3G network in case of
VoLTE deployed with ICS/SRVCC.

Table 3: Assessing the VoLTE approach

19
An upgrade of the HLR may prove indispensable to support the IP-SM-GW.
20
IP Short Message Gateway. Besides, the HSS may need an upgrade for IP-SM-GW support as well

19
2.3 VoLTE Logical Architecture

2.3.1 4G LTE network architecture

The VoLTE application relies on the same modern communication network as is the case
for the Rich Communications Suite (RCS), fixed broadband VoIP or messaging services
(few exceptional features may apply to the LTE network21). As a consequence, VoLTE
exploits the reliability of current commercial networks with regard to operations,
performance, scaling, etc. What is more, this reliability strengthens the VoLTE’s attraction
since it leverages the service convergence across the web and mobile world. Some
specific functions, which represent a service differentiation for the VoLTE operator, are
also provided by VoLTE:

 Emergency calling
 Smooth handover of in-progress calls from VoLTE to CS
 Intelligent Network (IN) service migration
 QoS

Having this in mind, MNOs should exploit the VoLTE’s (added) value by intensifying the
partnerships with application providers. In fact, VoLTE is conceived to ‘fit in’ within the
current subscriber’s voice experience22 by being transparent (whether the call is a VoLTE
or CS one) and simply dependent on the attached radio access technology. As will be
described later, novel wideband codecs may offer a superior voice quality and bring the
user experience to the next level.

Furthermore, the 3GPP Release 8 builds the foundation for the VoLTE profile. Obviously,
not all the functionalities stated by this release are necessary to guarantee VoLTE
compliance (some features rely on higher releases). Particularly, the VoLTE logical
architecture comprise:

 The Radio Access Network (RAN) is characterized by the Evolved Universal

Terrestrial Radio Access Network (E-UTRAN23)

21
to guarantee QoS or handover voice calls from LTE to 2G/3G
22
known from CS voice devices
23
e.g. LTE. In fact, LTE radio capabilities for FDD/TDD LTE apply to VoLTE

20
  The VoLTE UE comprises the features allowing the EPC and LTE RAN
accessibility (and this, in turn, mobile connectivity). Moreover, a VoLTE IMS
application as well as a nested IMS stack are necessary when granting VoLTE
services.

 The IMS Core Network delivers the service layer for the supply of Multimedia
Telephony (MMTel).

 The Evolved Packet Core (EPC) represents the Core Network.

The VoLTE architecture24 is depicted in Fig. 7: here, the blend of functional nodes into a
given physical node application appears possible (e.g. PGW and SGW). In this case, these
specific interfaces (S5 for example) become internal interfaces.

2.3.2 VoLTE Functional Node Description

The (VoLTE architecture’s) relevant functional nodes as stated by 3GPP are explained in
Fig. 725. Further information regarding the VoLTE Service Description and Implementation
Guidelines can be found in [31].

a) VoLTE User Equipment (UE)

This refers to the LTE capable equipment necessary to attach to the EPC, e.g. a UE using
the LTE-Uu interface to connect to the EPC

b) Evolved Universal Terrestrial Access Network (E-UTRAN)

EUTRAN comprises a unique node linking with the UE: the eNodeB, which houses the
Medium Access Control (MAC), Physical (PHY) and Radio Link Control (RLC)26 layers.
Moreover, the eNodeB is in charge of admission control, radio resource management,
(de)compression packet headers (DL/UL) or the enforcement of agreed QoS.

24
Interconnect and roaming are also included. In addition, the Gm (UE to P-CSCF) and Ut (UE to TAS) interfaces are built-in
in the VoLTE architecture. This is, however, not depicted in Fig. 7.
25
Further information available in 3GPP TS 23.002 [10]
26
these incorporate the features of encryption and user-plane header compression

21
Table 4 provides further information on the other two components, the Evolved Packet
Core (c) and IP Multimedia Subsystem (d). Particularly, IMS stands for the control structure
empowering the upcoming generation of IP multimedia services. Some of the omitted IMS
components27 in Table 4 are thoroughly explained in [31].

27
Interconnection Border Control Function/Transition Gateway (IBCF/TrGW), IMS-ALG/IMS-AGW (IMS Application Level
Gateway/IMS Access Gateway), Media Resource Function (MRF), MGCF/IMS-MGW (Media Gateway Control Function / IMS
Media Gateway)

22
Figure 7: VoLTE Logical Architecture [31] S

23
 c) Evolved Packet Core (EPC) d) IP Multimedia Subsystem (IMS)
Home Subscriber Server (HSS): Besides the storage of subscribers’ dynamic and static Telephony Application Server (TAS): This is an IMS Application Server supporting a
information, the HSS delivers user profile data to the IMS core and MME while the IMS registration series of binding Multimedia Telephony (MMTel) services28.
or UE attachment is performed.
Serving Gateway (SGW): The SGW characterizes the local mobility anchor for inter-eNodeB Proxy Call Session Control Function (P-CSCF): The IMS-enabled VoLTE UE firstly
handovers and inter-3GPP mobility29. Apart from packet forwarding and routing, the SGW contacts the P-CSCF for session signalling. By delivering SIP messages among the IMS
replicates the user traffic given a legitimate interception. Core Network and the UE, the P-CSCF performs the role of a SIP proxy. Furthermore, it
maintains security links between the VoLTE UE and itself.
Policy Charging and Rules Function (PCRF): Basically, the PCRF cares for an appropriate user Serving Call Session Control Function (S-CSCF): The S-CSCF supervises and registers
plane traffic handling in compliance with the subscriber’s defined settings. Also, it is responsible for specific sessions intended for billing. It also performs session set-up/control, routing jobs as
flow-based charging supervision as well as policy control arbitration. In this sense, the PCRF well as the role of a SIP registrar for VoLTE UE30. Additionally, .the S-CSCF administers
regulates the treatment for a service data flow when it comes to perform its enforcement role. calls comprising the end-user and HSS after the successful registration.
Packet Data Network (PDN) Gateway: The connectivity among external PDNs (Internet, IMS) Interrogating Call Session Control Function (I-CSCF): For all links having a subscriber
and the UE is ensured by the PDN Gateway (PGW): for instance, real-time connectivity to several of the operator's network as destination, the I-CSCF has to be connected. In the course of
PDNs may apply. The provision of a UE traffic entry/exit point as well as the packet the IMS registration, the I-CSCF queries the HSS to find the proper S-CSCF to handle the
screening/filtering, policy enforcement or charging support represent some tasks belonging to the (registration) request31.
PGW.
Mobility Management Entity (MME): For the LTE access network, the MME characterizes its BGCF (Breakout Gateway Control Function): Defining the following hop for the
main control node: as such it cares along with the HSS for the user authentication and allows the conveyance of SIP messages, based on the routing setup information, is managed by the
UE to camp on the service provider’s PMN. The MME deals also with the bearer (de)activation BGCF. There are, essentially, two termination events in (peer) IMS domains and CS ones:
procedure and handles the generation/allocation of provisional identities to the UE. in the first case, the BGCF pick out the suitable Interconnection Border Control
Function/Transition Gateway (IBCF) to manage the liaison to the peer IMS network. In the
Likewise, the MME delivers the control plane functionality to enable mobility among 2G/3G and second case, the BGCF takes the proper MGCF based on expected CS domain breakout.
LTE access networks. Last but not least, it is accountable for selecting the Serving Gateway MGCF/IMS-MGW (Media Gateway Control Function / IMS Media Gateway): This
(throughout the initial attachment and intra-LTE handover) for the UE. component is in charge of the control/media plane interworking32, which also comprises CS
networks founded on the Bearer Independent Call Control/ ISDN User Part (BICC/ISUP)

Table 4: VoLTE architecture: EPC and IMS components

28
Stated by 3GPP and outlined within the GSMA PRD IR.92 [29]. It involves supplementary service functionalities for example
29
by handing on the traffic to the PGW and 2G/3G
30
which the I-CSCF and HSS previously assigned to it
31
Upon call termination, the S-CSCF to which the subscriber is registered on is queried.
32
at the network interconnect point to CS domains
24
2.4 Assessing the Voice Quality Performance

Providing voice services over IP-based mobile networks requires not only telecom features
(e.g. QoS) but also end-to-end proven network-to-device functionality. Definitely, voice
represents an essential service and has, therefore, become a key distinguishing feature in
view of the acute competition within the telecom industry. And it is here where High
Definition (HD) voice33 comes into play: based on wideband audio links, the human voice
can be replicated more precisely, leading to an indeed natural sounding speech.

Certainly, traditional telephony is limited by traditional standards which results in

subscribers perceiving a considerable quality enhancement with HD Voice. For instance,
even the current digital telephony standards34 are built on 1960s digital circuit technology
and 1930s microphone know-how. G.711 was (prior to the introduction of HD voice) the
voice quality standard, meaning that mobile telephony was able to provide less than this
quality level.

Fig. 8 depicts the criteria determining the voice quality experienced: the handset cares for
many of these parameters, others are however dependent on the network capabilities35. In
fact, the voice quality among the PS (VoLTE enabled in LTE) and CS networks remains
the same since an identical voice codec mode is applied. It is the delay parameter that
makes the difference due to the substantial difficulty in managing it.

Particularly for PS services, delay stands for the crucial feature when it comes to attain an
ideal balance in terms of voice quality versus voice capacity (mainly on the radio network).
In this sense, the jitter buffer manager’s performance becomes decisive in facilitating
outstanding VoLTE services (Fig. 8) as a great amount of (packet delay) variations on the
radio network have to be managed36.

33
codec G.722.2
34
ITU-T G.711 Pulse Code Modulation (PCM)
35
and others are present in the handset, but influenced by the network
36
up to 80msec per link has to be handled according to 3GPP TS 23.203 [4]

25
 Figure 8: Parameters determining the user-perceived voice quality [28]

2.4.1 Leveraging the Operators’ Assets

Operators should pay special attention to exploit the high quality potential telecom voice
services are able to offer. In effect, carrier-grade VoIP services founded on VoLTE are at
all times prioritized in the mobile network compared to other data traffic. Over the core and
radio networks, QoS mechanisms are in place to care for key real-time services and their
quality. Over-The-Top (OTT) VoIP services run, on the contrary, as best-effort37 data over
the mobile broadband access. And here is where the codecs for HD voice play an
important role: they allow MNOs to employ them (e.g. codecs) with no major side-effects
on the radio frequency (RF) capacity. Moreover, operators are able to combine a certain
level of HD voice even when moving to 2G/3G. This efficiency advantage means that the
new codec for HD VoLTE will neither demand any improvements of the VoLTE radio
infrastructure nor additional RF capacity.

37
e.g. no guarantee of good voice service quality for users

26
Global interoperability is also a relevant operator asset to build on. Indeed, the global
E.16438 community builds a distinctive entity for billions of phone users: to reach one
person’s handset, wherever in the world, through the phone number is a valuable asset
that telephony service providers have built over the years. In this sense, operators can
benefit of a lead position over OTT opponents by further delivering interoperable voice
services. Still, advancing the voice quality experience even further, mainly based on
interoperable HD voice enhancements, will unleash superior advantages to all
stakeholders.

Beyond HD Voice
For users to experience HD voice via diverse IP-based voice applications, services are
compelled to apply interoperable voice codecs. Many of them are, nevertheless, being
developed throughout the market, further intensifying the voice services’ fragmentation and
rising transcoding expenses. As previously mentioned, the advantages of telecom-grade
services might become obsolete if no interoperability is guaranteed. Nevertheless, VoLTE
relies on the HD voice codec as its handset’s default (AMR-NB, AMR-WB). What is more,
some steps beyond HD voice have already been undertaken.

These efforts led to Evolved HD voice for LTE (EVS), which has been standardized in
3GPP and included in the profile IR.92 [29] and implies that HD voice services will offer a
more natural voice quality in any sort of communication setting. This is accomplished by
including a broader audio range for the perceptible frequency band as shown in Fig. 9.

Based on the IMS-based (3GPP) VoLTE, no additional components are necessary for
EVS. In other words, those network elements having an integrated media handling
functionality (e.g. SBG39, BGF and MRF) may need an upgrade. For the case of VoLTE
systems founded in IR.92 [29], just the addition of the EVS codec to the control plane
would be required. From the subscriber’s perspective, the UE has to fulfil some
requirements (acoustics, processing capability, codec support) to fully enjoy EVS. Detailed
information on this matter can be found in [24].

38
E.164 represents an ITU-T recommendation defining a numbering plan for the worldwide public switched telephone
network (PSTN) and a few other data networks
39
Session Border Gateway (SBG), Border Gateway Function (BGF) and Media Resource Function (MRF)

27
 Figure 9: Audio bandwidths for mobile voice services. [24]

2.4.2 Development of audio testing

Mobile handsets and VoIP networks make use not only of codecs that considerably modify
the (original) sound, but also apply limited bandwidth/bitrate. Regardless of these
limitations, the perceived voice (quality) is still satisfactory as a result of the right tradeoff
among various audio metrics such as the frequency response and a satisfactory speech
perception.

For the purpose of voice quality measurement, the Perceptual Evaluation of Speech
Quality (PESQ) was developed in the year 2000 and rapidly became an extensively
applied tool. Its successor, POLQA (Perceptual Objective Listening Quality Assessment),
additionally supports LTE, VoIP technologies, HD Voice, etc. Similar to PESQ, POLQA
assessments are closely related to tests with human subjects.

28
POLQA40 has been recognized as the modern Mean Opinion Score (MOS) benchmarking
technology for mobile networks. The horizontal axis in Fig. 10 depicts the evolution of
network technologies existing at the time of their development. A novel POLQA version
was implemented in the third quarter 201441.

Figure 10: Evolution of ITU-T recommendations for voice quality testing [37]

POLQA and PESQ reproduce the expected (measurement) results of humans with actual
speech samples, thus enabling speech-delivery system designers to guarantee a top
subscribers’ satisfaction level (refer to Table 5). Furthermore, POLQA is designed for these
supplementary test tasks:

 Discontinuous Transmission (DTX), Comfort Noise Insertion

 Terminal testing: Analysis of the influence of the transducer during
transmission/reception
 Voice Enhancement Devices (VED), Noise Reduction (NR)
 Voice Activity Detection (VAD), Automatic Gain Control (AGC)

40
the third generation perceptual voice quality test method standardized as P.863 by the ITU-T in 2011
41
The POLQA Coalition (OPTICOM, SwissQual and TNO) proposed an evolved version to ITU-T Study Group 12, which
was approved as Rec. P.863 Edition 2.4 in 09/2014.

29
 PESQ POLQA
Correct scoring with high background noise No Yes
Measurements with acoustic transducers No Yes
Effects of speech level in samples No Yes
AMR vs. EVRC codec comparison No Yes
Narrowband (300 Hz - 3400 Hz) Yes Yes
Wideband (100 Hz - 7000 Hz) Yes Yes
Super Wideband (50 Hz - 14000 Hz) No Yes
Linear Frequency distortion sensitivity No Yes

Table 5: PESQ versus POLQA. Adapted from [19]

2.4.3 PESQ/POLQA methodology

The audio quality of various (speech) samples was traditionally evaluated by several
persons within the range 1 to 5 (5 being the best score). All weighted single scores
generated then the Mean Opinion Score (MOS). However, such measurements relying on
people were time consuming and costly.

To address this issue, tools that comprise algorithms relying on psychoacoustic patterns
(e.g. PESQ, POLQA) can be applied. Ideally, these patterns would be consistent with the
findings obtained from human groups. Indeed, this algorithm-based approach permits fast
reiterations throughout the development process. Interestingly, the reproducible
(measurement) results are regarded as unbiased since they are independent of any test
condition or human behavior.

30
 Figure 11: Perceptual audio test block diagram [19]

Fig. 11 depicts how perceptual audio examinations work: both signals, the degraded and
the reference one, are independently level-aligned and screened with the receiving
handset’s transmission features. Moreover, both signals are time-aligned so as to balance
minor time alterations arising from coding, delay or jitter. Next, a conversion from the time-
amplitude into a frequency-loudness domain (e.g. auditory translation) takes place for the
two aligned/filtered signals. This can be interpreted as the alterations noticeable by a
human hearer.

In a later step, both signal representations are deducted from each other, yielding an
estimation of the perceptible variations: these (variations) are then gathered over time and
individually weighted based on various distortion sources42. Finally, the Mean Opinion
Score (MOS) value is derived to label the voice quality (Table 6).

MOS Quality Impairment

5 Excellent Imperceptible
4 Good Perceptible but not annoying
3 Fair Slightly annoying
2 Poor Annoying
1 Bad Very annoying

Table 6: MOS Score Summary

42
e.g. subject to added distortions or if portions of the signals were absent after the broadcast (drop-outs).

31
3 VoLTE Parameter Optimization

Deploying VoLTE requires a number of optimization steps (in the radio and transport
network) to gain the full benefit of the technology’s potential. Ideally, the success rate and
the retainability of a VoLTE call must exceed the level provided by CS connections.

Basically, network optimization involves the activation of features and optimization of

parameters such as:

 Robust Header Compression reduces the bandwidth associated with the headers
used to transport relatively small encoded audio packets
 Transmission Time Interval (TTI) Bundling overcomes the limitation of using short
(1ms) TTIs at cell boundaries

These features contribute to make a VoLTE call reliable while still providing more efficiency
compared to OTT VoIP applications. Special attention is paid, therefore, to decreasing the
necessary bandwidth for voice and get the most out of the capacity. By doing so, the
handover, setup and call completion success rates could be further improved.

In terms of handset power consumption, the focus lies on the one hand on the
enhancement of the handset architecture: the integration of VoLTE into the chip set or the
sleep mode available to the application processor fall into this category. On the other side,
radio features like the Discontinuous reception (DRX) intended to conserve the UE’s
battery life during a VoLTE conversation complement the (handset) power saving
solutions. Other valuable drivers are:

 Dedicated Bearers allow for the prioritization of VoLTE audio packets over all other
best-effort traffic
 Semi-Persistent Scheduling (SPS) reduces the complexity and overhead of the
continuous allocation of DL/UL physical layer resource blocks to transport the audio
traffic

Additional (optimization) features in the field of speech quality, depending primarily on the
voice codec sampling rate and the derived audio bandwidth, will be discussed separately.

32
3.1 Robust Header Compression (RoHC)

RoHC takes advantages of the redundancy of information present in

 the headers of subsequent packets in the same audio stream and

 various headers in different protocol layers

with the goal of decreasing the header size used to transport VoLTE audio. In fact, the 40-
60 bytes of header length can be condensed to 3-4 bytes. And this applies to VoLTE calls,
which usually comprise small encoded audio packets transmitted every 20ms.

Particularly, the size of the data is smaller than the headers for the protocols used to carry
the encoded data. For VoLTE deployments based on IPv6, the mix of RTP/UDP/IP
headers can sum up to 40-60 bytes long headers. This means that a VoLTE encoded
audio transmission (relying on the Wideband-AMR43 codec and RoHC) is reduced from 75
bytes to approximately 35 bytes.

As shown in Fig. 12, RoHC is used over the air interface to preserve the RAN’s valuable
bandwidth. In fact, RoHC may accomplish around 50% reduction in the size of VoLTE
audio transmissions: the required bandwidth for any call is reduced while the amount of
users on a given eNodeB site is increased.

Figure 12: RoHC Compression and Decompression at the UE and eNodeB [44]

43
Adaptive Multi-Rate Wideband (AMR-WB) codec G.722.2

33
3.2 Transmission Time Interval (TTI) Bundling

LTE presents a shorter TTI (1ms subframe) than earlier defined in cellular technologies. In
fact, a smaller TTI enables low over-the-air latency for real-time applications since
resource scheduling is performed for every TTI. Some uplink issues are, though, caused
by the short TTI in specific settings of an eNodeB’s coverage: given an UE located at a cell
edge (with deteriorating reception and no possibility of enhancing its transmission power),
the eNodeB may activate the TTI bundling. In other words, the UE will intensify the error
detection and correction related to each data broadcast by transmitting over several TTIs.
Based on this improved approach (e.g. error detection/correction), overall latency is
diminished compared to the use of a single TTI.

As presented in Fig. 13, TTI bundling contributes to lower-latency VoLTE data at cell
edges, where information errors are predictable. Instead of waiting for the HARQ44 process
(typically 8ms/period) to send a data retransmission request, a data retransmission is
assumed by TTI bundling: by doing so, several data packets are filled into a HARQ
interlace period in advance. As a result, every packet comprises the identical source data
coded with 4 distinct groups of error detection/correction bits.

Generally speaking, the short 1ms TTI (along with its bundling approach) represents for
VoLTE a considerable advantage: it improves the uplink performance at cell edges by
means of multiple bundled TTIs to transfer higher error detection and correction
information.

Figure 13: Effect of TTI bundling on latency [44]

44
Hybrid Automatic Repeat Request

34
3.3 Discontinuous Reception (DRX)

For packet-based voice services, each encoded audio packet transmission (usually 20ms
for VoLTE) is followed by a period of no transmission. DRX exploits, for instance, these
silent periods by switching off several components such as digital signal processors, UE’s
RF receiver or A/D converters. As a result, the device’s battery life can be conserved.

Nevertheless, setting a too long ‘sleep period’ might involve latency issues at the detriment
of the wanted QCI value level (e.g. a predefined performance profile). Particularly, the
network establishing the DRX pattern (and also being aware of the data downlink schedule
to the UE) should carefully select a suitable DRX configuration in view of likely
retransmission needs and latency constraints of the application. In this sense, special
attention should be paid to the influence of TTI on DRX in LTE so as to improve DRX
settings in favor of power saving.

As depicted in Fig. 14, the two DRX operation modes can be applied according to the
pause duration within a chat: a long DRX deactivates the UE receiver for a longer time
period in a situation where audio packets are not frequently delivered. The eNodeB’s MAC
Layer or an activity timer at the UE would regulate the transition from short to long DRX
and vice versa. Worth mentioning is the absence of DRX short or long cycles in UMTS
(only fixed-length ones are available).

Figure 14: Long and Short DRX [44]

35
3.4 Dedicated Bearers

Despite the increased spectral efficiency offered by LTE, over-the-air bandwidth remains a
limited and valuable resource. Moreover, Internet-based VoIP services usually generate a
greater load on mobile networks, leading to negative (quality) side-effects. Operators need,
therefore, to invest in more capacity to cope with the OTT voice traffic since (their) voice
applications naturally consume additional network resources. In other words, each
application along with the associated data strives for that finite bandwidth.

In fact, the encoded voice packets an off-the-shelf VoIP client produces do not show a
significant difference when compared with data traffic derived from web browsing or video
streaming (from a network’s perspective). As a result, the network tries to bundle all users’
(generic) data traffic into a single common channel45. Interestingly, with a default EPS
Bearer no control over the service quality is possible. This implies a ‘best effort’ approach
to transport all generic traffic to the Internet PDN. By the time the network resources are
(temporarily) exhausted, latency may greatly fluctuate and packets be dropped due to data
traffic queuing.

To overcome these drawbacks, which cannot be tolerated for real-time applications, LTE
introduces the EPS Dedicated Bearer: It permits the isolation of certain types of data traffic
(VoIP traffic versus FTP file download for instance). For VoLTE, an EPS dedicated Bearer
will exclusively convey encoded voice packets among the UE and an IMS PDN-GW. Also,
each Dedicated Bearer can possess diverse quality characteristics defining a QoS Class
Identifier (QCI).

 Packet Delay Budget: The Maximum tolerable end-to-end delay among the UE and
the PDN-GW
 Resource Type: Guaranteed Bit Rate (GBR) and Non-GBR
 Allocation Retention Priority: In case the capacity is depleted, scheduling is based
on the value assigned (‘1’ equals to the highest level)
 Packet Error Loss Rate: Threshold level of IP packets not effectively received by
the Packet Data Convergence Protocol (PDCP)

45
e.g. the Physical Downlink and Physical Uplink Shared Channels PDSCH/PUSCH, having one or more EPS bearers
linking the UE to the PDN-GW

36
 Table 7 further presents standardized QCI values, ‘1’ being assigned to VoLTE traffic that
would be prioritized over all best-effort traffic on the Default Bearer.

QCI Resource Priority Packet Delay Packet Error Example Services

Type Budget (ms) Loss Rate
1 GBR 2 100 10-2 Conversational Voice
-3
2 GBR 4 150 10 Conversational Video (live streaming)
3 GBR 5 300 10-6 Non-conversational video (buffered
streaming)
4 GBR 3 50 10-3 Real-time gaming
-6
5 Non- GBR 1 100 10 IMS Signalling
-3
6 Non- GBR 7 100 10 Voice, video (live streaming), interactive
gaming
7 Non- GBR 6 300 10-6 Video (buffered streaming)
8 Non- GBR 8 300 10-6 Video (buffered streaming)TCP-based
9 Non- GBR 9 300 10 -6 (email, www, FTP)

Table 7: Standardized QCI Values. Adapted from [44]

3.5 Semi-Persistent Scheduling (SPS)

As previously stated, the physical layer relies on common channels (PDSCH/PUSCH) to
deliver the data contained in the logical bearers. A way to allocate these channels appears
necessary to avoid several users (on an eNodeB) attempting to use identical resources at
the same time. Specifically, a LTE carrier is split into several subcarriers within the
frequency domain, whereas in the time domain every subcarrier is gathered into 0.5ms
time slots.

A cluster of twelve subcarriers in one time slot is referred to as a Resource Block (RB) and
is defined as the smallest portion of the LTE physical layer (resource) that can be placed to
a UE (refer to Fig. 15). Moreover, VoLTE encounters a challenge when it comes to
granting control channel overhead: as each downlink/uplink RB must be granted, the
(control channel) overhead for the ongoing allocation of RBs becomes unmanageable.

37
 Figure 15: LTE Resource Block (RB) in the time and frequency domain (Physical Layer) [44]

To overcome this, Semi-Persistent Scheduling (SPS) was introduced to reduce granting

overhead for applications such as VoLTE. In effect, the steady transmission configuration
of VoLTE packets allows SPS to make a continuing grant of RBs instead of separately
arranging each download/uplink RB. Of course, this involves a Radio Resource Control
(RRC) message used to determine the RB grant’s periodicity.

The green boxes in Fig. 16 illustrate the SPS-arranged RBs for a VoLTE call. As further
shown by the orange box, additional RBs can be dynamically organized for data traffic
while SPS is enabled (for example for a file download during a VoLTE call).

Figure 16: Semi-Persistent Scheduling [44]

38
Important to consider are the pauses (e.g. short-time silence periods) in the course of a
VoLTE conversation: this may represent a drawback since (physical layer) resources get
lost in case the SPS grant is continued. In this case, the expiration of the SPS grant after
some network-defined transmission slots would be recommendable (uplink). In the end, it’s
all about finding the right balance between efficiency improvement (for shared data
channels) and minimizing control channel overhead.

39
 4 VoLTE Testing

4.1 Primary Implications

Table 8 depicts the relevant VoLTE testing areas and their related measures.

Functional Aspects Audio Quality Performance Features

Verify parameters based on Signal analysis: Comparison of Assess the effect of fading, IP
optimal settings the reference signal with the one impairments and noise on audio
Goal

obtained quality. Note: different test settings

for different test purposes
Check the LTE attach order Audio quality assessment with Analyze the UE capability to deal
proven approaches (e.g. with IP traffic deficiencies
PESQ/POLQA)

Validate the IMS registration. Verify the UE audio quality for a Assess the impact of fading on a
Further test whether the UE is VoLTE call VoLTE call sound quality
capable of registering to the IMS
Server
Confirm that the Session Acoustic signal assessment Test setup: CMW + UPV Audio
Measures

Initiation Protocol (SIP) can based on 3GPP TS 26.132 [4] analyzer + AMU
launch a voice session Release 10

Check if the UE is able to Test setup: artificial mouth and

perform a VoLTE call ear are required along with an
audio isolated chamber

Examine the basic audio

functionality through simple
echo confirmation in loopback
mode

Table 8: VoLTE Testing Areas. Based on [39], [40]

Within the testing context, some key indicators have to be considered:

 Accessibility: It primarily assess the network access in order to get a service. A

typical indicator would be a (network attach) failure ratio lower than 1%.

40
  Retainability: This term refers to the service capability to last (e.g. keep on
providing the service) as long as this is requested by the subscriber. In terms of
voice services, retainability sheds some light on network performance, especially
when addressing the question how well can the network maintain calls (from the
setup to their usual end). This data is vital in identifying poor performing cells, thus
preserving the network against dropped calls. Here, a typical indicator is given by a
dropped call ratio lower than 2%.

 Integrity: This indicates whether the network observed sound quality is regarded
as ‘good’ in more than 95% of the total trials. A typical indicator of the speech
quality would be characterized by 95% of all samples having a MOS better than
3.0.

4.1.1 Voice Quality of VoLTE versus OTT Voice Services

A voice quality evaluation of VoLTE and few others OTT voice services was conducted by
Ericsson over commercial VoLTE-enabled LTE networks in South Korea [25] [28]. For this
purpose, measurements were performed on two operator networks, employing three
different first-class LTE smartphone brands: one of them was not available for VoLTE and
another one was not permitted into one of the networks46.

The tests were undertaken throughout the day and night (e.g. during low network load
times). Moreover, a drive test was organized, generating around one hour of data per
network, service and mobile handset. Also for the testing duration, a drive test application
was linked to two identical handsets and the call nature was device-device. Furthermore,
to reproduce the noticeable voice quality throughout a conversation, two factors needed
were simultaneously examined: the speech path delay (SPD) as well as an estimation of
the one-way audio quality (MOS-LQOSWB47). Lastly, both parameters were depicted in
parallel to evaluate the conversation’s audio quality.

46
Hence the number of measurements (e.g. the quantity of square-dots) in Fig. 17 varies among the services evaluated
47
This represents the measurement unit used by POLQA on wideband speech

41
Fig. 17 illustrates the summarized outcomes of the obtained MOS-LQOSWB and SPD during
the busy hours. In general, each square-dot48 in the chart displays the 95th percentile SPD
scores (e.g. 95% of the scores are better=lower) and 5th percentile MOS-LQOSWB results
(meaning that 95% of the results are better=higher). Particularly, the dotted lines delimit a
gray highlighted zone indicating common objectives for a first-class VoLTE service relying
on both MOS-LQOSWB and SPD. Specifically, SPD will strive for less than 225ms and
MOS-LQOSWB should achieve higher than 3.5.

Figure 17: Voice quality results based on MOS-LQOSWB and SPD: VoLTE versus OTT voice
services [28]

As seen in Fig. 17, VoLTE surpassed those (minimum) marks and none of the OTT voice
services were even close. For VoLTE, the device as well as the network go through an
explicit validation/integration procedure: the aim being the verification of the handset’s jitter
buffer and its ability to deal with major packet delay deviations, mainly near the (LTE
network) cell border. On the contrary, a more volatile handset performance would be
experienced with OTT voice services: not only because of the missing capability in
handling (packet delay) variations, but also due to the missing QoS features.

48
based on the measured combination of network and device

42
Generally speaking, all services may accomplish comparable MOS scores, but then again
network priority is indispensable when a low SPD and hence an excellent voice quality is
required. Additionally, handsets are explicitly devised for the provision of telephony
services like VoLTE, since various standardization demands have to be met49. In a
nutshell, VoLTE as a real-time telephony service is given priority over other services and
handled via the MNO’s network end-to-end50. This is, however, not the case for OTT VoIP
services, which are combined with further data traffic and carried over the LTE network on
a best-effort basis.

Performance of Combined Voice and Data services over LTE

Despite the fact that HD voice represents a differentiator for VoLTE subscribers, others
factors such as the call setup time (CST) have to be considered as well: the CST can be
enhanced to be as fast as one second, characterizing a major progress to the usual 4s
goal in CS systems. Thanks to DRX, the handset is inactive for ‘sleeping periods’ during
the packet delivery and reception, leading to a minor energy usage for VoLTE calls.
Therefore, these (VoLTE) handsets are able to maintain longer conversations in
comparison to CS (WCDMA) or OTT calls.

During a VoLTE call, subscribers may continue enjoying high-speed LTE data. What is
more, data and voice services can cohabit in an efficient and seamless way based on the
LTE radio interface design. Particularly, VoLTE offers a considerable network capacity as
well as built-in QoS mechanisms, thus allowing for a reliable call quality irrespective of the
large data traffic load.

4.1.2 Mobile Network Test: Testing Framework

For LTE subscribers to fully exploit the advantages the more advanced RF technology
offers (including the data-intensive media services), a reliable network operational
performance has to be ensured. This implies ongoing infrastructure inspections to timely
identify and amend arising software and hardware issues.

49
this is not the case for OTT services
50
IMS, EPC, LTE with policy control

43
In the last few years, as the telecom sector has transitioned to HSPA and LTE
technologies, the network intelligence has been increasingly ‘relocated’ from the core to
the network’s edge and then into the E-UTRAN Node B (eNB)51. In other words, a good
deal of the control and decision making is currently implemented within the eNBs. For
instance, the communication taking place among base stations (BS) and the UE can be
best supervised by devices implicated in the particular transactions. For sure, one major
driver (for the shift of the decision making entity involved in traffic management) is the
latency demands for the sake of the LTE performance: the signaling control traffic must no
longer pass through several network nodes when there is a modification for a UE.

Organizing a Drive Testing (DT)

The extensive performance assessment of a wireless network infrastructure (and its
components) represents the fundamental goal of such a test. In fact, diverse settings like
the availability level, call quality or the network coverage fall within the DT scope.
Moreover, its execution appears meaningful after the installation of novel cell sites so as to
make sure that these are properly incorporated into the network. Likewise, tests can be
performed on a continuing basis to identify the root cause for a particular issue or to
assess the network operational efficiency. Undeniably, the introduction of LTE has raised
the significance of drive testing for MNOs.

Generally speaking, DT entails the data retrieval straight from a real network. The
predefinition of several Key Performance Indicators (KPI) and their post-test verification
would help analyze the QoS supplied by the network, thus analyzing the network
performance from the subscriber’s perspective. In this regard, potential problems like the
frequency of dropped calls within a specific cell site (prior and after its upgrading) due to
errors in the course of the implementation may be a good example for an issue a KPI need
to address.

51
also moving the traffic management toward the network boundary

44
Furthermore, DT is widely applied to supervise the handover52 spot between legacy and
LTE technologies. Particularly, the duration to conclude an originated handover, the data-
interruption time (e.g. while moving among technologies) for the end-user and the success
rates represent some KPIs that are carefully examined. Network optimization and
benchmarking belong to the drive testing scope since they support MNOs in assessing the
degree of (network) capacity utilization

Following resources are usually necessary during a DT:

 Measuring device: Typically, a minimum of two handsets will be employed: one is

active (so as to perform call-based measurements), whereas the second device
remains only attached to the network (e.g. inactive). Moreover, the call and idle
mode performance can be assessed at different locations throughout the route.

For instance, short and long calls are carried out to test diverse performance
settings: long calls may provide an insight on handover characteristics, while short
calls rather address signal-based features. Besides, the successful handover
execution among contiguous cells can be determined based on speech quality,
signal strength or download speed.

 GPS system: An illustration of the network coverage, including the likely areas of
concern, can be generated based on the obtained test and GPS data.

 Test vehicle: This will be used to travel a specific route while obtaining test data.
Obviously, the testing could also be performed on foot.

 Computer comprising a data analysis/acquisition software: Used for storage

and (real-time) analysis of the data obtained.

52
The signal quality, cell ID and neighbor information are evaluated before and after a handover for optimization purposes

45
4.2 Mobile Network Test 2014: Connect Magazine

As in previous years, the magazine ‘Connect’ assigned P3 communications with the

realization of the network benchmarks. For this purpose, two measurement vehicles were
deployed: each of them provided with ten53 Samsung Galaxy S4 LTE+ handsets. Wherever
applicable, the device’s firmware54 corresponded to the MNO’s original one. Another test
team, equipped with a portable measurement system, performed walk tests covering
indoors and trains.

4.2.1 Organizational aspects

At the beginning of 2014, an initial meeting (P3 communications and Connect) took place
to review the previous network test and outline the upcoming test framework: this
comprised not only the test locations within the DACH55 countries, but also a preliminary
schedule based on the smartphones to be employed.

Based on the testing principles, each MNO’s CTO56 received comprehensive information
concerning the testing logistics (e.g. services under test, locations, etc.) as well as the
timetable. A feedback was then requested to prevent the test to be performed during a
considerable network adjustment. In a later step, the presentation’s official version
including any comments on the test methodology was provided to the Technical
Management.

Moreover, additional data provided to the MNO comprised the test outcomes prior to the
final publication. Here, the aim was to offer adequate time to each MNO for the coming
announcement of the test results. For the scoring and measurement methodology,
Connect had at all times the final say.

53
six to eight devices used for testing subject to the country
54
Should no particular firmware be available, the up-to-date Samsung firmware was used
55
Germany, Austria and Switzerland
56
Chief Technology Officer

46
4.2.2 Test methodology
Throughout the measurements, two vehicles with predefined routes were in the same city,
however not at the same location57. For static measurements, the vehicles halted at
designated ‘zones of interest’ (train stations, airports or highly inhabited residential
spaces). Interestingly, it was here where the measurement systems competed for the
existing networks capacities with other subscribers.

In the case of Germany, the chosen locations were expected to account for a significant
population share. Consequently, tests were performed in 21 large German cities while the
smaller ones were travelled by one measurement vehicle on the way from one heavily
populated city to the other. Furthermore, 16 densely populated cities were included in the
case of Austria and Switzerland: In fact, around 20 small cities were also covered in each
country. Similar to the German setting, highways were mostly travelled by the
measurement vehicles.

57
to avoid disturbances from one car to the other

47
Some basic points in terms of test methodology are described in Table 9.

Test Methodology
Three main areas fell within the data test scope: data service capability, indoor
coverage and download performance based on the (static) ‘Kepler’ test website58. In
fact, using 3-4 Samsung handsets in the vehicles, many well-known websites were
visited employing the integrated browser (sites were previously chosen according to the
Smartphone Data

Alexa ranking59).

For indoor or in-vehicle coverage, an offset of 12dB was introduced in the way from the
antennas fixed on the vehicle’s roof to the phones (in the car). Particularly, two
antennas were used for each device since LTE applies the MIMO (Multiple Input
Multiple Output) technology.

Concerning the data service performance, upload and download files (1MB, 3MB) were
sent from or to a test server. What is more, high definition videos (HD, 720p, 11.9MB,
30s) were streamed with the player provided by the phones.
For 2014, all MNOs in the DACH region offered LTE tariffs: all phones were hence
preset to ‘LTE preferred’ mode and subject to CSFB after a call was initiated (previous
to the call: attachment to a LTE cell).
Smartphone
Telephony

Basically, the tests on voice and data were performed using different handsets. For
realistic conditions, data traffic was passing through the handsets for the duration of a
call. Moreover, telephony was examined from one vehicle to the other (e.g. mobile-to-
mobile). The obtained speech samples were then evaluated based on the POLQA
algorithm.
For the walk tests, the procedure applied to data and voice services remained valid.
trains/indoors
Walk tests: in

This implied measurement spots while using the public transport or being around public
buildings (including train stations, museums or airport terminals). Wherever possible,
the test team travelled by train to other cities to assess the performance on the way to
their destination.

Table 9: Network Test: Methodology

58
as defined by ETSI for these testing scope
59
A ranking system set by alexa.com (a subsidiary of amazon.com) that basically audits and makes public the frequency of
visits on various Web sites.

48
4.2.3 Germany: Telephony Test Results
The German telecom sector in 2014 was not only characterized by considerable
investments in LTE or fiber-optic networks, but also by takeovers (Kabel Deutschland by
Vodafone) as well as mergers (E-Plus/Telefónica). Furthermore, the revenues in this
60
German industry fell by 0,9% to EUR 65 billion, which implies that the significant profit
reduction derived from voice services could not be balanced by the (minor) growth in data
services. Also, the earnings from mobile service decreased by approximately 1,6% to
around EUR 19 billion (versus prior year’s period). This can be understood as the result of
the widely usage of IP services and social networks and, ultimately, to other MNO’s
aggressive pricing strategies. And despite these adverse developments, MNOs continue
modernizing their networks. Some of the key findings for the voice services under test can
be summarized as follows:

 The achieved success rates of Vodafone and Deutsche Telekom remained with
98,5% at a high level (refer to Table 10). For their part, E-Plus and O2 are left
behind with rates near 96%
 Telekom, O2 and Vodafone possess a LTE share in the range 80%-90% in large
cities, causing numerous CSFB handovers
 Concerning the drive test: outdoor measurements for Deutsche Telekom and
Vodafone stayed at the same (good) level as the previous network test (a reliability
enhancement of around 4,4% was even observed).
 The measurements on railways shed some negative light on the overall voice
performance: E-Plus unsuccessful calls summed up to 21%, O2 and Telekom show
only a slightly better ratio of 13% lost calls. For instance, around 3% of the calls
carried some noise.
 In terms of CSFB performance, some HD voice parameters were missing
generating a poor speech quality (the case of O2)

As stated in the first chapter, the numerical data throughout this paper uses a comma "," as
decimal mark.

60
according to an assessment of BITKOM (Federal Association for Information Technology, Telecommunications and New
Media)

49
 Germany ‐ Telephony
Operators
Deutsche 
Vodafone O2 E‐Plus
Parameter Telekom
TELEPHONY  Call Success Ratio (%) 98,5 98,5 95,7 96,0

(BIG CITIES  Call Setup Time (s) 6,3 6,3 9,1 6,3

Speech Quality (MOS‐LQO) 3,3 3,4 2,6 3,3
OUTDOOR)
Call Sustainability (%) 99,6 99,6 99,2 98,8
Call Success Ratio (%) 98,0 98,3 96,1 97,4
TELEPHONY 
Call Setup Time (s) 6,3 6,0 9,2 5,9
(BIG CITIES 
Speech Quality (MOS‐LQO) 3,4 3,4 2,6 3,4
INDOOR)
Call Sustainability (%) 99,7 99,6 99,8 98,7
Call Success Ratio (%) 97,6 95,4 93,1 92,5
TELEPHONY  Call Setup Time (s) 7,1 7,1 10,2 6,9
(HIGHWAYS) Speech Quality (MOS‐LQO) 3,1 3,0 2,6 3,0
Call Sustainability (%) 98,9 99,2 98,9 98,1
Call Success Ratio (%) 86,7 83,9 85,5 78,6
TELEPHONY  Call Setup Time (s) 7,6 7,3 9,7 8,0
(TRAINS) Speech Quality (MOS‐LQO) 2,7 2,7 2,5 2,6
Call Sustainability (%) 96,5 97,9 96,5 96,5

Table 10: Telephony (D): detailed test results according to operator and location. Adapted from [48]

50
4.2.4 Germany: Data Test Results
Indeed excellent service ratios for large cities and drive testing (better than 99% in almost
all cases) were obtained. Particularly, Deutsche Telekom benefits from a satisfactory LTE
coverage (1800 MHz). Some highlights are described below:

 The transfer/download speed, despite the rising traffic volume since the previous
network test, witnessed a sound level especially for Deutsche Telekom (up to 40
Mbit/s and more than 11 Mbit/s in 90% of all measurements). Moreover, further
indoor measurements attested the outdoor performance accurately.
 The upload tests (indoors) indicated, on the contrary, a slightly underperformance.
 A very good reliability was provided for the city outdoor setting. Streaming HD
videos was successfully accomplished by most operators (E-Plus took the lead
over O2 and Vodafone)
 Just one operator delivered poor results in regard to file retrieval and static internet
surfing (E-Plus: 98,6% and 98,4% respectively).
 Adequate coverage was not always given for small towns: Vodafone ended up
behind Telekom in terms of file transfer and internet surfing performance, which
pointed to the reduced speed and reliability.

51
 Germany ‐ Data
Operators
Deutsche 
Parameter Vodafone O2 E‐Plus
Telekom
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,8 / 99,9 99,9 / 99,7 99,7 / 99,0 99,4 / 98,4
Avg. Session Time (s/s) 3,2 / 0,8 3,8 / 1,0 4,0 / 1,1 5,1 / 2,8
DATA  FILE‐DOWNLOAD (3 MB)
(BIG CITIES  Success Ratio/Avg. Session Time (%/s) 99,8 / 1,6 99,8 / 2,5 99,9 / 2,6 98,6 / 4,5
OUTDOOR) 90 % faster than (kbit/s) 11734 5475 5428 3340
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,5 / 0,9 98,7 / 1,4 98,9 / 1,5 96,8 / 2,1
Video playouts without interruptions (%) 98,9 97 98,2 94,3
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,4 / 99,6 99,4 / 99,7 99,2 / 99,5 98,9 / 98,8
Avg. Session Time (s/s) 3,5 / 0,8 4,4 / 1,1 5,3 / 1,3 5,5 / 1,7
DATA  FILE‐DOWNLOAD (3 MB)
(BIG CITIES  Success Ratio/Avg. Session Time (%/s) 99,6 / 2,1 99,7 / 4,3 99,2 / 4,5 98,0 / 3,9
INDOOR) 90 % faster than (kbit/s) 10729 4006 3023 3450
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,7 / 0,9 97,3 / 1,5 96,2 / 2,1 96,9 / 2,1
Video playouts without interruptions (%) 99,3 93,9 91,4 94,4
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,6 / 99,7 98,9 / 96,8 98,6 / 98,0 99,3 / 97,8
Avg. Session Time (s/s) 3,4 / 0,9 4,6 / 1,9 5,5 / 2,2 6,1 / 3,3
DATA 
FILE‐DOWNLOAD (3 MB)
(SMALL 
Success Ratio/Avg. Session Time (%/s) 99,5 / 4,1 98,4 / 4,8 99,4 / 8,2 98,9 / 6,3
CITIES 
90 % faster than (kbit/s) 8386 2975 1956 2626
OUTDOOR)
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,4 / 1,1 97,2 / 2,1 91,3 / 2,6 95,5 / 3,0
Video playouts without interruptions (%) 97,7 90,9 88,4 89,8
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,1 / 99,4 98,3 / 98,0 93,3 / 91,3 94,5 / 91,4
Avg. Session Time (s/s) 3,8 / 1,5 4,4 / 1,5 5,8 / 2,4 6,8 / 4,1
DATA  FILE‐DOWNLOAD (3 MB)
ON  Success Ratio/Avg. Session Time (%/s) 98,4 / 3,5 98,2 / 4,6 91 / 12,2 92,1 / 13,1
HIGHWAYS 90 % faster than (kbit/s) 5572 3836 269 593
YOUTUBE HD
Success Ratio / Start Time (%/s) 98,1 / 1,4 96,1 / 1,8 94,6 / 2,1 89,9 / 3,1
Video playouts without interruptions (%) 96,8 92,4 91,8 85,1
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 89,0 / 90,6 83,6 / 82,1 80,9 / 72,7 87,6 / 83,9
Avg. Session Time (s/s) 5,7 / 1,1 7,4 / 2,2 8,1 / 2,6 7,9 / 2,4
FILE‐DOWNLOAD (3 MB)
DATA 
Success Ratio/Avg. Session Time (%/s) 86,1 / 7,1 72,8 / 18,9 61,9 / 23,9 77,6 / 12,8
ON TRAINS
90 % faster than (kbit/s) 320 111 100 144
YOUTUBE HD
Success Ratio / Start Time (%/s) 93,4 / 1,4 87,5 / 3,6 90,2 / 4,4 85,3 / 4,0
Video playouts without interruptions (%) 92,7 76,8 74,4 78,9

Table 11: Data (D): detailed test results according to operator and location. Adapted from [48]

52
 Mobile Network Testing 2014 ‐ Germany
100% Telephony Data

90%

80%

70%

60%

50%

40%
Big Cities Big Cities Highways Trains Big Cities Big Cities Small Cities Highways Trains
(outdoor) (indoor) (outdoor) (indoor)
Deutsche Telekom Vodafone O2 E‐Plus

Figure 18: Germany: detailed test results according to operator and location. Based on [48]

4.2.5 Austria: Telephony Test Results

The telecom landscape in Austria is, with over 13 million SIM cards by the end of 2013,
primarily concentrated on the mobile networks. The underlying data volume, which
increased by 57% from 2012 to 2013, attests for the unbroken capacity development
operators are expected to cope with. Furthermore, the solid (market) status demonstrated
in past years is attributable the local MNO’s high performance levels. Furthermore, some
key factors identified in the network test 2014 are:

 For 2014, the achievement level of telephony in Austria remained at a high level. In
other words, high performing German operators represent a baseline for Austrian
counterparts
 T-Mobile and Drei succeeded in deploying LTE, which is translated in success
ratios in the range 98,4%-98,7% in cities. These results were clearly surpassed by
A1 Telekom Austria and Drei
 In terms of stability on the highways among cities: A1 Telekom Austria and Drei
accomplished, despite the slightly decrease in performance, success ratios around
97%-98%, followed by T-Mobile with 95%

53
 Austria ‐ Telephony
Operators
A1 Telekom 
Parameter T‐Mobile Drei
Austria
Call Success Ratio (%) 98,7 / 99,1 98,4 / 98,7 98,7 / 99,8
TELEPHONY 
Call Setup Time (s) 5,7 / 6 5,7 / 5,6 6,8 / 6,9
(BIG CITIES 
Speech Quality (MOS‐LQO) 3,6 / 3,7 3,5 / 3,6 3,7 / 3,8
OUTDOOR/INDOOR)
Call Sustainability (%) 99,9 / 99,7 99,6 / 99,7 99,8 / 99,7
Call Success Ratio (%) 97,9 / 90,8 95,0 / 85,9 97,1 / 82,6
TELEPHONY  Call Setup Time (s) 5,7 / 5,5 6,5 / 6,3 7,4 / 7,7
(HIGHWAYS/TRAINS) Speech Quality (MOS‐LQO) 3,6 / 3,4 3,3 / 3,2 3,5 / 3,4
Call Sustainability (%) 99,5 / 98,7 99,4 / 98,7 99,6 / 98,5

Table 12: Telephony (A): detailed test results according to operator and location. Adapted from [48]

4.2.6 Austria: Data Test Results

Data services entails, for many operators, a (market) differentiation potential. As a result,
impressive success ratios (better than 99,5%61) were realized in large cities for
uploads/downloads as well as web surfing. In the same manner, MNOs showed a similar
development in small cities: T-Mobile and Drei are trailing A1 Telekom Austria in terms of
performance (even if reliability is moderately higher for T-Mobile and Drei). Some of the
key findings can be summarized as follows:

 For A1, the average download speed advanced from 13 Mbit/s to 52 Mbit/s62. This
improvement is, however, not reflected when surfing in the web: the loading time of
2.2 seconds for A1 is just 0,8 seconds faster than for the two other MNOs
 The average data rates for Drei improved, based on 2013 figures, from 7 Mbit/s to
30 Mbit/s
 The data performance on trains showed a clear ranking winner (A1), followed by
Drei and T-Mobile. Moreover, the case of highways resemble to that of cities, with
Drei closing the gap to A1’s performance and outperforming T-Mobile.

61
only for YouTube HD two MNO surpassed the 99% threshold due to the marginally rise of interruptions
62
this is associated with the (YouTube) HD service offered by A1

54
 Austria ‐ Data
Operators
A1 Telekom 
Parameter T‐Mobile Drei
Austria
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,7 / 99,6 99,8 / 99,7 99,8 / 99,5
Avg. Session Time (s/s) 2,2 / 0,6 2,8 / 1,2 3,0 / 1,6
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
(BIG CITIES  Success Ratio/Avg. Session Time (%/s) 99,5 / 99,7 99,9 / 99,8 99,5 / 99,5
OUTDOOR) 90 % faster than (kbit/s) 2355 / 4379 7397 / 1136 4419 / 1207
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,9 / 100 99,8 / 98,6 99,8 / 99
Video playouts without interruptions (%) 99,9 / 99,8 99,6 / 97,8 99,7 / 97,1
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,7 / 99,8 99,7 / 99,4 99,8 / 100
Avg. Session Time (s/s) 2,5 / 0,6 2,8 / 0,7 3,8 / 1,5
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
(BIG CITIES  Success Ratio/Avg. Session Time (%/s) 99,8 / 99,8 99,8 / 99,6 99,2 / 100
INDOOR) 90 % faster than (kbit/s) 2132 / 2782 11452 / 1159 3266 / 634
YOUTUBE HD
Success Ratio / Start Time (%/s) 100 / 99,8 99,8 / 98,8 99,6 / 96,0
Video playouts without interruptions (%) 100 / 99,6 99,8 / 98,6 99,2 / 91,8
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 97,9 / 96,9 99,0 / 97,7 99,8 / 99,6
Avg. Session Time (s/s) 3,0 / 1,8 3,7 / 2,0 3,7 / 2,2
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
(SMALL CITIES  Success Ratio/Avg. Session Time (%/s) 99,2 / 98,4 100 / 99,2 98,5 / 99,3
OUTDOOR) 90 % faster than (kbit/s) 1085 / 890 5063 / 667 2421 / 1239
YOUTUBE HD
Success Ratio / Start Time (%/s) 100 / 98,4 98,4 / 98,4 100 / 94,6
Video playouts without interruptions (%) 100 / 97,6 98,4 / 97,6 100 / 90,0
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 97,7 / 97,3 98,0 / 96,5 98,9 / 98,4
Avg. Session Time (s/s) 3,4 / 2,2 4,8 / 3,2 4,2 / 2,8
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
ON  Success Ratio/Avg. Session Time (%/s) 97,5 / 98,8 97,8 / 99,1 98,0 / 99,1
HIGHWAYS 90 % faster than (kbit/s) 841 / 530 1207 / 318 2482 / 863
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,1 / 98,3 97,1 / 94,5 98,4 / 96,1
Video playouts without interruptions (%) 98,9 / 98,1 96,4 / 93,3 97,3 / 93,9
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 96,1 / 96,4 96,5 / 96,9 96,8 / 97,8
Avg. Session Time (s/s) 4,1 / 1,8 5,5 / 2,5 6,2 / 3,0
FILE‐DOWNLOAD (3 MB / 1 MB)
DATA 
Success Ratio/Avg. Session Time (%/s) 91,4 / 95,8 95,3 / 98,4 91,6 / 97,2
ON TRAINS
90 % faster than (kbit/s) 392 / 301 763 / 317 827 / 195
YOUTUBE HD
Success Ratio / Start Time (%/s) 98,3 / 94,6 93,9 / 84,1 97,1 / 80,7
Video playouts without interruptions (%) 97,8 / 90,4 91,7 / 81,8 91,9 / 69,3

Table 13: Data (A): detailed test results according to operator and location. Adapted from [48]

55
 Mobile Network Testing 2014 ‐ Austria
100% Telephony Data

90%

80%

70%

60%

50%

40%
Big Cities Big Cities Highways Trains Big Cities Big Cities Small Highways Trains
(outdoor) (indoor) (outdoor) (indoor) Cities
A1 Telekom Austria T‐Mobile Drei
Figure 19: Austria: detailed test results according to operator and location. Based on [48]

4.2.7 Switzerland: Telephony Test Results

The Swiss telecom market, having three operators by the end of 2014, is characterized by
a wide range of voice/data services as well as its innovative spirit. For instance, the
increasing merging of IT, telecommunications and media is shaping the market, thus
attracting additional global players. For their part, these novel contestants are providing
cost-effective Internet-based services like SMS messaging, telephony or TV. In this regard,
cloud-based approaches are becoming sort of ‘game changers’ (e.g. more relevant) as
processing power, storage capacity, etc. tend to progressively migrate to the Internet.

As a result of the high market penetration, the growth of SIM cards ownership (0,8% by the
end of 2014) was for a consecutive time sluggish. Particularly, the (achievement) gap
railways versus highways became evident again in spite of the exceptionally high railway
usage in this country. Some highlights are described below:

 All three MNOs achieved success ratios in the range of 99% (telephony)
 In terms of call setup time, Sunrise outperformed (despite the similarity of the high
LTE share) all other MNOs with 4,4 – 4,5 seconds. As for speech quality, Sunrise
and Swisscom attained good MOS levels (3,4 – 3,5), tightly followed by Orange

56
  As for reliability on the highways among cities, the achieved level for all MNOs
stayed indeed high. Likewise, the calls quality on trains was remarkably good
(Orange being the top performer in this category).

Switzerland ‐ Telephony
Operators

Parameter Swisscom Sunrise Orange

Call Success Ratio (%) 98,6 / 99,4 98,9 / 99,4 98,4 / 99,6

TELEPHONY 
Call Setup Time (s) 7,5 / 7,2 4,5 / 4,4 5,9 / 5,8
(BIG CITIES 
Speech Quality (MOS‐LQO) 3,4 / 3,5 3,4 / 3,5 3,2 / 3,4
OUTDOOR/INDOOR)
Call Sustainability (%) 99,7 / 99,5 99,4 / 99,8 99,5 / 99,9
Call Success Ratio (%) 97,8 / 93,7 97,7 / 91,6 97,2 / 96,5
TELEPHONY Call Setup Time (s) 7,5 / 7,9 4,6 / 4,9 5,8 / 5,9
 (HIGHWAYS/TRAINS) Speech Quality (MOS‐LQO) 3,4 / 3,3 3,3 / 3,2 3,1 / 3,1
Call Sustainability (%) 99,7 / 98,8 98,8 / 97,2 99,6 / 99,5

Table 14:Telephony (CH): detailed test results according to operator and location. Adapted from [48]

4.2.8 Switzerland: Data Test Results

In fact, average speeds showed an enhancement of more than 50% compared to 2014,
leading to almost perfectly retrieved YouTube videos. This can be translated in average
upload speeds of 27Mbit/s and 35Mbit/s for downloads (Swisscom). Or in other words:
retrieving a web site did last 3 seconds at maximum.

 Success ratios of 99%-100% for mobile Internet in cities are common for all MNOs
 The (reliability) performance on trains and highways ended up slightly behind the
100% mark. For YouTube HD, this gap was marginally higher (especially for
Orange).

57
 Switzerland ‐ Data
Operators

Parameter Swisscom Sunrise Orange

WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,4 / 99,6 99,5 / 99,6 99,4 / 99,2
Avg. Session Time (s/s) 2,7 / 1,1 2,6 / 1,1 3,0 / 1,1
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
(BIG CITIES  Success Ratio/Avg. Session Time (%/s) 99,8 / 99,9 99,8 / 99,8 99,5 / 99,9
OUTDOOR) 90 % faster than (kbit/s) 11619 / 4613 4908 / 1871 7221 / 2497
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,9 / 99,3 99,7 / 98,7 99,8 / 99,0
Video playouts without interruptions (%) 99,8 / 98,9 99,5 / 97,3 99,7 / 98,4
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,5 / 100 99,5 / 99,8 99,6 / 100
Avg. Session Time (s/s) 2,8 / 0,9 2,6 / 1,0 3,0 / 0,9
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
(BIG CITIES  Success Ratio/Avg. Session Time (%/s) 99,8 / 99,7 99,8 / 99,5 99,8 / 100
INDOOR) 90 % faster than (kbit/s) 10774 / 3023 8742 / 2700 10150 / 3963
YOUTUBE HD
Success Ratio / Start Time (%/s) 100 / 99,8 100 / 99,3 100 / 99,8
Video playouts without interruptions (%) 100 / 99,8 100 / 99,3 100 / 99,6
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,9 / 100 99,8 / 99,7 99,9 / 99,7
Avg. Session Time (s/s) 2,7 / 1,1 2,7 / 1,4 2,9 / 1,1
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
(SMALL CITIES  Success Ratio/Avg. Session Time (%/s) 100 / 100 100 / 100 100 / 100
OUTDOOR) 90 % faster than (kbit/s) 13296 / 6182 4286 / 1399 6289 / 1254
YOUTUBE HD
Success Ratio / Start Time (%/s) 100 / 100 100 / 96,1 100 / 99,3
Video playouts without interruptions (%) 100 / 98,7 100 / 94,8 100 / 98,0
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 99,1 / 98,9 98,7 / 99,3 97,9 / 96,3
Avg. Session Time (s/s) 3,1 / 1,6 3,2 / 1,9 3,6 / 1,8
DATA  FILE‐DOWNLOAD (3 MB / 1 MB)
ON  Success Ratio/Avg. Session Time (%/s) 99,3 / 99,5 98,0 / 99,0 96,8 / 97,4
HIGHWAYS 90 % faster than (kbit/s) 6379 / 2059 2535 / 1094 2537 / 1376
YOUTUBE HD
Success Ratio / Start Time (%/s) 99,7 / 98,2 98,8 / 96,1 99,1 / 96,2
Video playouts without interruptions (%) 99,7 / 97,2 98,2 / 94,2 98,8 / 95,4
WEB‐PAGE DOWNLOAD (LIVE/STATIC)
Success Ratio (%/%) 98,2 / 98,1 97,9 / 98,0 97,6 / 98,0
Avg. Session Time (s/s) 3,8 / 1,3 4,4 / 1,9 4,0 / 1,5
FILE‐DOWNLOAD (3 MB / 1 MB)
DATA 
Success Ratio/Avg. Session Time (%/s) 98,6 / 97,6 95,6 / 96,6 97,0 / 99,0
ON TRAINS
90 % faster than (kbit/s) 2685 / 692 1622 / 928 2352 / 1165
YOUTUBE HD
Success Ratio / Start Time (%/s) 98,5 / 96,4 97,0 / 89,3 98,0 / 96,4
Video playouts without interruptions (%) 98,5 / 93,8 95,5 / 82,7 97,5 / 93,8

Table 15: Data (CH): detailed test results according to operator and location. Adapted from [48]

58
 Mobile Network Testing 2014 ‐ Switzerland
100% Telephony Data

90%

80%

70%

60%

50%

40%
Big Cities Big Cities Highways Trains Big Cities Big Cities Small Highways Trains
(outdoor) (indoor) (outdoor) (indoor) Cities
Swisscom Sunrise Orange

Figure 20: Switzerland: detailed test results according to operator and location. Based on [48]

4.2.9 Mobile Network Test: Overall Results DACH Region

Undeniably, providing fast LTE speeds combined with reliable voice (2G/3G) involves, as
exposed by the last mobile network benchmark, a balancing act for each MNO. Taking the
top performing Swiss operator (Swisscom) as a reference point, it is clearly demonstrated
how the room for development could be best capitalized. Swisscom63 offered even in rural
areas an excellent coverage level.

In Austria, the performance benchmark is definitely given by A1 Telekom Austria with its
high-speed and wide-ranging data network. With regard to (high quality) speech
transmission, A1 Telekom Austria also outperformed other competitors. Nevertheless, a
network performance improvement was observed for T-Mobile: from 13 Mbit/s to 28 Mbit/s
soared the download throughput64. By contrast, T-Mobile’s some weaker points comprised
HD videos and data uploads in small cities or telephony on railways. Drei’s data services
were likewise negatively rated during the test.

63
with 60% market share
64
average within the past 12 months (previous to the network test)

59
Last but not least, the German Deutsche Telekom (local test winner) was able to provide
high quality telephony65 services in large and less populated cities. In addition, O2 showed
some voice quality weaknesses owing to its recent LTE deployment. For the merged
Telefónica/E-Plus, the upcoming network test should show a progress in terms of LTE
coverage and ultimately of their data service level. Vodafone shone, for its part, with fast
and reliable data services.

Table 16 shows the summarized results for the network benchmark in the DACH region.

Table 16: DACH Region: Telephony and Data test results by operator. Adapted from [48]

4.2.10 Austria: VoLTE Friendly User Tests

Some operators66 across the DACH region participated during 2014 in Friendly User Tests
(FUT): this assessment was, however, limited to specific (country) regions activated for
VoLTE measurements. In fact, the MNOs provided also the smartphones used for the
measurements (equipped with their own VoLTE firmware).

As shown in the anonymized Table 17, the failure rates for four network operators were
extremely low. The two remaining ones have major (E) or minor (A) issues to solve before
seriously thinking of implementing VoLTE. If the connections with at least a handover to
the 3G network are considered, the success rate for three operators reached the 99%

65
this strong performance was also observed on rural routes despite some weaknesses in trains
66
Deutsche Telekom, Vodafone, E-Plus, O2, Swisscom and A1 Telekom Austria

60
mark. Moreover, the short call setup time are noticeable with 2 to maximum 4 seconds on
average (with precondition). Though, a source of debate may be given by the 0,5 seconds
(call setup time) achieved by operator D.

In addition, P3 communications has examined the latency in voice transmissions67: three

operators achieved below 150 milliseconds, whereas the rest touched the limit of
approximately 200 milliseconds (a range required for an undisturbed conversation).

Concerning voice quality, the accomplished MOS value (3,3 - 3,9) improved by 0,8 points
compared to the test average for 2013. This significant progress proves VoLTE as a
suitable standard in terms of reliability and quality. Remarkably, VoLTE cannot only
support additional call partners within a radio cell, but also (thanks to the all-IP approach)
makes the core network cost-effective for operators.

Anonymized VoLTE call statistics DACH
Operators

Parameter A B C D E F

Call Success Ratio (%) 95,8 100,0 99,3 99,4 87,6 100,0

Call Setup Time (s) 2,0 2,1 2,1 2,1 3,7 1,9
Call Setup Time min./max.(s) 0,8/10,5 1,0/5,5 0,6/4,7 0,5/6,0 2,5/19,1 1,1/5,7
Speech Quality (MOS‐LQO) 3,5 3,8 3,6 3,8 3,3 3,9

Table 17: First VoLTE call statistics DACH Region 09/2014. Adapted from [49]

4.3 Mobile Network Test 2014: CHIP Magazine

The network test 2014 of the magazine ‘CHIP’ focused solely on four German operators.
The test engineers of NET CHECK, CHIP’s test partner, covered some 5.600 kilometers
(one third of the time by car, public transport and on foot respectively) and collected more
than 100.000 individual measurements. For the most densely populated cities in each
state, the testers assigned enough time to allow for detailed measurements.

67
defined as the time required for words to be heard by the opposite party

61
The evaluation scheme emphasized the mobile data access (60%) over telephony (40%).
Furthermore, different modules for telephony and data access performance were defined:
measurements along the highways were weighted with 10% and those along the state
roads and public metropolitan/suburban railway with 25% each. The remaining 40% were
equally assigned to measurements within the main traffic arteries and those obtained while
walking.

4.3.1 Test methodology

Two measuring systems consisting of twelve operational smartphones were employed:
four Galaxy S3 LTE were dedicated to test the telephony, other four S3 LTE retrieved
permanently videos from YouTube and four Samsung Galaxy S5 were testing the
performance when uploading/downloading large amounts of data. Also, the Galaxy S5
already support LTE Category 4 with multiple antenna systems (MIMO), allowing
theoretical data rates of up to 150 Mbit/s.

The test vehicle travelled throughout German highways and rural roads, driving them on
secondary routes through sparsely populated areas. While driving, the equipment
comprised two measuring systems (QualiPoc Free Rider II), each having 6 smartphones.
The smartphones were kept in a specially designed box on the car roof. No vehicle-
mounted external antennas were used as in 2013. Instead, the internal antennas of
smartphones were used to test the multi-antenna MIMO technology under realistic
conditions.

While using the public transport, the measuring systems were moved from the roof box into
two special backpacks. The testers then walked into the inner cities and travelled among
the main traffic arteries of each city to guarantee the widest possible test coverage.

4.3.2 Germany: Telephony Test Results

For the assessment of mobile networks, calls (to the smartphones) were initiated from fixed
line networks as well as from the mobile devices themselves. A successful call setup was
planned to last around two minutes, in which several speech samples were exchanged
between the smartphone and the measurement computer (at the other end). The
underlying speech sample was automatically evaluated based on the POLQA algorithm.

62
 As a rule of thumb, samples showing a MOS value below 2,7 can cause serious problems
in understanding the other party. The evaluation scheme for voice quality involves,
therefore, the success rate (25%), call setup time (10%), call stability (25%) and the
quantity of ‘poor’ speech samples (40%) with a MOS score below 2,7.

Germany ‐ Telephony
Operators
Deutsche 
Parameter Vodafone O2 E‐Plus
Telekom
Call Success Ratio (%) 97,8 95,7 96,9 98,0

TELEPHONY  Interrupted Calls (%) 0,0 1,0 3,0 3,3

(HIGHWAYS) Call Setup Time (s) 7,0 7,6 8,0 6,5

Share Poor Speech Samples ‐ 
4,9 4,6 8,9 6,5
MOS below 2,7 (%)
Call Success Ratio (%) 98,3 95,9 96,6 96,6
TELEPHONY  Interrupted Calls (%) 1,0 1,8 1,7 1,7
(COUNTRY  Call Setup Time (s) 6,8 7,3 7,7 6,6
ROADS) Share Poor Speech Samples ‐ 
4,0 4,4 5,1 7,6
MOS below 2,7 (%)
Call Success Ratio (%) 98,5 98,9 97,0 98,9
Interrupted Calls (%) 0,3 0,3 0,3 1,1
CITIES 
Call Setup Time (s) 6,7 7,6 8,7 7,5
(BY CAR)
Share Poor Speech Samples ‐ 
2,5 2,4 4,0 5,4
MOS below 2,7 (%)
Call Success Ratio (%) 97,9 97,2 97,5 97,6
CITIES  Interrupted Calls (%) 2,0 2,4 1,8 2,4
(BY PUBLIC  Call Setup Time (s) 6,7 7,5 8,2 7,1
TRANSPORT) Share Poor Speech Samples ‐ 
2,7 2,7 4,7 3,4
MOS below 2,7 (%)
Call Success Ratio (%) 98,4 98,0 97,7 97,2
Interrupted Calls (%) 0,0 0,0 0,5 0,2
CITIES 
Call Setup Time (s) 7,0 7,6 8,7 7,3
(ON FOOT)
Share Poor Speech Samples ‐ 
2,2 1,5 3,5 1,4
MOS below 2,7 (%)

Table 18: Telephony (D): detailed test results according to location. Adapted from [20]

63
4.3.3 Germany: Data Test Results
As stated above, eight smartphones68 were dedicated to evaluate the mobile data access
(download/upload rates), streaming of YouTube videos and web browsing. The predefined
weighting ratios represented 60% for web browsing, 25% for data throughput and 15% for
video streaming.

Moreover, the evaluation of the data throughput foresaw additional (performance) factors
like the maximum retrieved data within 30 seconds: the success rate accounted for 50%,
the average data throughput over the measuring time (25%) and the percentage of data
connections having a data throughput above 2 Mbit/s (upload) or 5 Mbit/s (download) were
weighted with 25% as well.

Besides, the reliability of web browsing was likewise severe evaluated: in case a test
website could not be completely displayed after 30 seconds, the test case was graded
unsuccessful. In this case, 50% of the points were assigned based on the success rate,
the remaining 50% awarded for the fastest possible website load. A similar approach
applied to streaming (YouTube) videos: 40% of the points were awarded for a successfully
streamed media file. Another 40% for brief interruptions and the last 20% were granted for
the shortest possible initial delay.

68
four Samsung Galaxy S5 and four Samsung Galaxy S3 LTE were used in parallel

64
 Germany ‐ Data
Operators
Deutsche 
Parameter Vodafone O2 E‐Plus
Telekom
Download Success Ratio (%) 99,1 93,0 87,0 92,0
Average Download Rate (Mbit/s) 21,3 13,6 10,8 8,2
Upload Success Ratio (%) 99,4 94,9 88,9 89,2
DATA  Average Upload Rate (Mbit/s) 9,9 10,4 5,4 2,5
(HIGHWAYS) Success Ratio Website retrieval (%) 95,7 93,7 76,7 85,4
Duration Website retrieval (s) 5,4 5,1 7,8 8,4
Success Ratio retrieved Youtube 
videos (%) 96,3 80,4 46,2 62,4
Download Success Ratio (%) 98,3 94,3 88,8 85,9
Average Download Rate (Mbit/s) 24,3 12,7 10,8 11,3
Upload Success Ratio (%) 98,3 93,3 87,0 85,7
DATA 
Average Upload Rate (Mbit/s) 13,7 8,9 6,3 4,1
(COUNTRY 
Success Ratio Website retrieval (%) 94,6 90,8 82,0 81,4
ROADS)
Duration Website retrieval (s) 5,7 6,3 6,9 7,8
Success Ratio retrieved Youtube 
videos (%) 87,8 80,9 54,5 60,9
Download Success Ratio (%) 99,0 95,9 99,0 98,0
Average Download Rate (Mbit/s) 37,3 17,5 13,6 14,7
Upload Success Ratio (%) 99,1 96,3 97,0 99,3
CITIES  Average Upload Rate (Mbit/s) 25,0 13,6 9,2 6,3
(BY CAR) Success Ratio Website retrieval (%) 98,6 98,5 97,5 97,7
Duration Website retrieval (s) 4,3 4,5 5,7 6,4
Success Ratio retrieved Youtube 
videos (%) 99,6 98,6 92,4 95,0
Download Success Ratio (%) 96,4 93,9 95,1 94,1
Average Download Rate (Mbit/s) 27,3 13,2 12,2 12,7
Upload Success Ratio (%) 96,0 93,3 93,4 93,4
CITIES 
14,4 9,2 6,2 4,2
(BY PUBLIC  Average Upload Rate (Mbit/s)
TRANSPORT) Success Ratio Website retrieval (%) 94,5 94,3 93,4 92,6
Duration Website retrieval (s) 5,3 5,2 6,3 6,9
Success Ratio retrieved Youtube 
videos (%) 95,3 86,3 89,9 83,4
Download Success Ratio (%) 99,8 94,6 99,3 99,3
Average Download Rate (Mbit/s) 33,1 18,1 14,2 15,9
Upload Success Ratio (%) 100,0 95,8 97,7 99,3
CITIES  Average Upload Rate (Mbit/s) 22,9 13,3 9,1 6,5
(ON FOOT) Success Ratio Website retrieval (%) 99,1 98,2 98,2 97,4
Duration Website retrieval (s) 4,3 4,3 5,4 6,2
Success Ratio retrieved Youtube 
videos (%) 99,6 94,7 97,7 90,9
Table 19: Data (D): detailed test results according to location. Adapted from [20]

65
5 Conclusion

According to [30] [47], the global ICT market rose by 4% in 2014. Particularly, this was the
result of the solid demand for telecom services and facilities, mainly in the USA and Asia.
Another highlight in the previous year was given by the strong (telecom) industry’s
consolidation pressure in Europe, triggered by dropping returns [30]. In the end, the
decisive objective of further investing in network roll-outs is to effectively deal with the
escalating data speed/volume requirements. Fortunately, MNOs are successfully showing
their capabilities in coping with these challenges. And this benefits not only subscribers or
the telecom sector, but also fosters a self-governing digital economy.

In this sense, VoLTE may boost a MNO’s network productivity and further enrich the
subscriber’s experience in terms of call setup times and speech quality. Consequently,
VoLTE embodies the MNO’s imminent strategy and progress towards novel
communication services this industry must deal with. For instance, it is not a matter of
deciding whether to implement VoLTE or not, but instead of selecting the proper tactical
approach to do it. In this sense, IMS’s cost-effectiveness and flexibility as well as its
capability to provide for IP services (regardless of the access method) enabled it to
become the favorite approach for delivering core services within an all-IP setting.

Furthermore, major efforts should be undertaken to deal with events threatening a first-
class voice experience, e.g. being outside LTE coverage (prior to a conversation) or
quitting the LTE coverage range after a call origination. This scenario is indeed relevant
since extensive LTE coverage, throughout the initial LTE deployments, cannot be
assumed.

Meanwhile, IP communications are being accepted as the next evolutionary step of core
mobile services. Operators are, therefore, called to continue investing in their networks
(e.g. IP infrastructure) so as to enjoy a superior brand positioning and be able to ‘switch-off’
their legacy networks in the medium term69. Additionally, greater control over
supplementary access methods (Wi-Fi for example) as well as an increased need for
MNO’s core applications is the VoLTE promise in the long run.

69
Or at least include legacy networks in an ICS architecture

66
Forthcoming challenges
The network test 2015 focused on new features like LTE Advanced along with Carrier
Aggregation (CA), (super) wideband voice connections as well as VoLTE measurements
[49]. A1 Telekom Austria successfully assessed its telephony service over LTE (VoLTE
trial) in the second quarter 2014, following the implementation of LTE CA in certain
Austrian cities by the end of 2014 (speeds of up to 300 Mbps were achieved with proper
devices). There is, however, a long way to go in view of the well-equipped and motivated
market players in the DACH region (Swisscom for instance).

Interoperability poses, in addition, a challenge when it comes to fully realize the inherent
potential of IP communications. In fact, the extent to which operators support
interconnection (and roaming to some degree) will decide on the MNO’s ability to exploit
the benefits coupled with IP communications.

Looking ahead
Despite some deployment issues or extensive consensus, the telecom sector gives a good
impression of being prepared for IP-communications. In this regard, the ‘journey’ away
from industry-driven towards to demand driven growth is expected as a direct
consequence of the emergent acceptance of HD calling and VoLTE in the subscriber’s
mind [30]. Undoubtedly, VoLTE is gradually being perceived as a premium device’s in-built
feature, likely to be found already ‘in the box’.

Finally, no matter where competitive pressures or regulation initiatives of technology

enhancements might take the telecom business within the next few years, MNO’s
(financial) endeavors in favor of VoLTE will enable them to handle in a forward-looking and
customer-friendly way.

67
Bibliography
[1] 3GPP, The Evolved Packet Core, available on http://www.3gpp.org/technologies/keywords-
acronyms/100-the-evolved-packet-core [last access on 01.02.2015]

[2] 3GPP, TR 23.856, Single Radio Voice Call Continuity (SR-VCC) enhancements; Stage 2.,
available on http://www.3gpp.org/ftp/Specs/html-info/23856.htm [last access on 01.03.2015]

[3] 3GPP. TS 22.278, Service requirements for the Evolved Packet System (EPS), available on
http://www.3gpp.org/ftp/Specs/html-info/22278.htm [last access on 01.03.2015]

[4] 3GPP. TS 23.216, Single Radio Voice Call Continuity; Stage 2, available on
http://www.3gpp.org/ftp/Specs/html-info/23216.htm [last access on 01.03.2015]

[5] 3GPP. TS 23.237, IP Multimedia Subsystem (IMS) Service Continuity; Stage 2, available on
http://www.3gpp.org/ftp/Specs/html-info/23237.htm [last access on 01.03.2015]

[6] 3GPP. TS 23.401, General Packet Radio Service (GPRS) enhancements for Evolved
Universal Terrestrial Radio Access Network (eUTRAN) access, available on
http://www.3gpp.org/ftp/Specs/html-info/23401.htm [last access on 01.03.2015]

[7] 3GPP, TS 24.237, IP Multimedia (IM) Core Network (CN) subsystem IP Multimedia
Subsystem (IMS) service continuity; Stage 3, available on
http://www.3gpp.org/ftp/Specs/html-info/24237.htm [last access on 01.03.2015]

[8] 3GPP, TS 25.304, User Equipment (UE) procedures in idle mode and procedures for cell
reselection in connected mode, available on http://www.3gpp.org/ftp/Specs/html-
info/25304.htm [last access on 01.03.2015]

[9] 3GPP. TS 22.173, IP Multimedia Core Network Subsystem (IMS) Multimedia Telephony
Service and supplementary services; Stage 1, available on
http://www.3gpp.org/DynaReport/22173.htm [last access on 01.03.2015]

[10] 3GPP, TS 23.292: IP Multimedia Subsystem (IMS) centralized services; Stage 2, available
on http://www.3gpp.org/DynaReport/23292.htm [last access on 01.03.2015]

[11] 3GPP, TS 24.008: Mobile radio interface Layer 3 specification; Core network protocols;
Stage 3, available on http://www.3gpp.org/dynareport/24008.htm [last access on
01.03.2015]

[12] 3GPP TS 23.228: IP Multimedia Subsystem (IMS); Stage 2, available on

http://www.3gpp.org/DynaReport/23228.htm [last access on 01.03.2015]

[13] 4G Americas, VoLTE and RCS Technology Evolution & Ecosystem, 2014

[14] Alcatel-Lucent, Options for Providing Voice over LTE and Their Impact on the GSM/UMTS
Network, Strategic White Paper, 2009

68
[15] Alcatel-Lucent, Service Consistency for Today’s VoLTE Subscribers, Technology White
Paper, 2011

[16] Alcatel-Lucent, Voice over LTE: The new mobile voice, Strategic White Paper, 2012

[17] Alcatel-Lucent, Service Continuity for Today’s VoLTE Subscribers, Technology White
Paper, 2012

[18] A. Perez, Voice over LTE: EPS and IMS Networks, John Wiley & Sons, First Edition, 2013

[19] Audio Precision, PESQ and POLQA Options, available on

http://www.ap.com/products/apx/perceptual [last access on 14.02.2015]

[20] CHIP Digital, Netztest 2014, available on http://www.chip.de/artikel/Der-haerteste-Handy-

Netztest-Deutschlands-Telekom-Vodafone-O2-und-E-Plus-im-Test_63944005.html [last
access on 01.03.2015]

[21] Cisco, Voice over Long Term Evolution Migration Strategies, White Paper, 2012

[22] Deutsche Telekom, Annual Report 2014, available on https://www.telekom.com/ar-2014

[last access on 10.04.2015]

[23] European Commission, Digital Agenda for Europe, Europe 2020 Strategy, available on
https://ec.europa.eu/digital-agenda/our-goals/pillar-i-digital-single-market [last access on
10.04.2015]

[24] Ericsson, Evolved HD voice for LTE, White Paper, 2014

[25] Ericsson, Mobility Report 2014, available on

http://www.ericsson.com/res/docs/2014/ericsson-mobility-report-november-2014.pdf [last
access on 19.03.2015]

[26] Ericsson, Mobility Report. Mobile World Congress Edition, available on

http://www.ericsson.com/res/docs/2015/ericsson-mobility-report-feb-2015-interim.pdf [last
access on 25.03.2015]

[27] Ericsson, Validating voice over LTE end-to-end, Ericsson Review, 2012

[28] Ericsson, Voice and video calling over LTE, White Paper, 2014

[29] GSMA, IMS Profile for Voice and SMS, Official Document IR.92, 2014

[30] GSMA, The Mobile Economy 2015, available on

http://www.gsmamobileeconomy.com/GSMA_Global_Mobile_Economy_Report_2015.pdf
[last access on 02.04.2015]

[31] GSMA, VoLTE Service Description and Implementation Guidelines, Official Document
FCM.01, 2014

[32] Huawei, Full HD Voice, White Paper, 2014

69
[33] J. Rankin, A. Costaiche, J. Zeto, Validating VoLTE - A Definitive Guide to Successful
Deployments, First Edition, CreateSpace Independent Publishing Platform, 2013

[34] M. Poikselkä, H. Holma, J. Hongisto, J. Kallio, A. Toskala, Voice over LTE (VoLTE), John
Wiley & Sons, First Edition, 2012

[35] Nokia Networks, Evolve to richer voice with Voice over LTE (VoLTE), White Paper, 2014

[36] Nokia Networks, Voice over LTE (VoLTE) Optimization, White Paper, 2014

[37] POLQA Coalition, POLQA Technology, available on http://polqa.info/index.html [last access

on 15.02.2015]

[38] Qualcomm, VoLTE with SRVCC: The second phase of voice evolution for mobile LTE
devices, White Paper, 2012

[39] Rohde & Schwarz, Next-Generation (3G/4G) Voice Quality Testing with POLQA, White
Paper, 2012

[40] Rohde & Schwarz, IOT & VoLTE, R&S IOT/PQA Seminar, available on
http://www.digitimes.com.tw/tw/b2b/seminar/service/download/053a111160/053a111160_ku
4os9wcy1p3lzfl8trd.pdf [last access on 25.02.2015]

[41] Signals Research Group, Behind the VoLTE curtain (Part 1), Vol. 10, Number 7 PREVIEW,
2014

[42] Signals Research Group, Behind the VoLTE curtain (Part 2), 2015

[43] Spirent, IMS Architecture. The LTE User Equipment Perspective, White Paper, 2014

[44] Spirent, VoLTE Deployment and the Radio Access Network. The LTE User Equipment
Perspective, White Paper, 2012

[45] Swisscom, Annual Report 2014, available on

https://www.swisscom.ch/en/about/investors/reports.html [last access on 15.04.2015]

[46] Telecoms.com Intelligence, VoLTE: Why, When and How?, White Paper, 2013

[47] Telekom Austria, Annual Report 2014, available on

http://www.telekomaustria.com/en/ir/annual-reports [last access on 09.04.2015]

[48] Weka Media Publishing, Connect Netztest 2014/2015, available on

http://http://www.connect.de/netztest/ [last access on 01.03.2015]

[49] Weka Media Publishing, Anruf der vierten Generation, Connect Magazine 12/2014, pages
70-74

70
List of Figures

Figure 1: Circuit and packet domains. LTE within the Evolved Packet System (EPS) [1] .... 7 
Figure 2: VoLTE’s strategic value for operators. Adapted from [14] ................................... 10 
Figure 3: Call setup time measurements in comparison [36] .............................................. 11 
Figure 4: Voice Strategies: Providing voice over LTE [35] .................................................. 14 
Figure 5: Common SRVCC mechanism [17] ...................................................................... 17 
Figure 6: Service continuity compared to service consistency [15]..................................... 18 
Figure 7: VoLTE Logical Architecture [31] S ....................................................................... 23 
Figure 8: Parameters determining the user-perceived voice quality [28] ............................ 26 
Figure 9: Audio bandwidths for mobile voice services. [24] ................................................ 28 
Figure 10: Evolution of ITU-T recommendations for voice quality testing [37] .................... 29 
Figure 11: Perceptual audio test block diagram [19] ........................................................... 31 
Figure 12: RoHC Compression and Decompression at the UE and eNodeB [44] .............. 33 
Figure 13: Effect of TTI bundling on latency [44] ................................................................ 34 
Figure 14: Long and Short DRX [44] ................................................................................... 35 
Figure 15: LTE Resource Block (RB) in the time and frequency domain (Physical Layer)
[44] ...................................................................................................................................... 38 
Figure 16: Semi-Persistent Scheduling [44] ........................................................................ 38 
Figure 17: Voice quality results based on MOS-LQOSWB and SPD: VoLTE versus OTT
voice services [28] .............................................................................................................. 42 
Figure 18: Germany: detailed test results according to operator and location. Based on [48]
............................................................................................................................................ 53 
Figure 19: Austria: detailed test results according to operator and location. Based on [48]56 
Figure 20: Switzerland: detailed test results according to operator and location. Based on
[48] ...................................................................................................................................... 59 

71
List of Tables

Table 1: VoLTE compared to other approaches for voice in LTE devices. Adapted from [16]
............................................................................................................................................ 13 
Table 2: Assessing the CSFB approach ............................................................................. 15 
Table 3: Assessing the VoLTE approach ............................................................................ 19 
Table 4: VoLTE architecture: EPC and IMS components ................................................... 24 
Table 5: PESQ versus POLQA. Adapted from [19]............................................................. 30 
Table 6: MOS Score Summary ........................................................................................... 31 
Table 7: Standardized QCI Values. Adapted from [44] ....................................................... 37 
Table 8: VoLTE Testing Areas. Based on [39], [40] ............................................................ 40 
Table 9: Network Test: Methodology .................................................................................. 48 
Table 10: Telephony (D): detailed test results according to operator and location. Adapted
from [48] .............................................................................................................................. 50 
Table 11: Data (D): detailed test results according to operator and location. Adapted from
[48] ...................................................................................................................................... 52 
Table 12: Telephony (A): detailed test results according to operator and location. Adapted
from [48] .............................................................................................................................. 54 
Table 13: Data (A): detailed test results according to operator and location. Adapted from
[48] ...................................................................................................................................... 55 
Table 14:Telephony (CH): detailed test results according to operator and location. Adapted
from [48] .............................................................................................................................. 57 
Table 15: Data (CH): detailed test results according to operator and location. Adapted from
[48] ...................................................................................................................................... 58 
Table 16: DACH Region: Telephony and Data test results by operator. Adapted from [48]
............................................................................................................................................ 60 
Table 17: First VoLTE call statistics DACH Region 09/2014. Adapted from [49] ................ 61 
Table 18: Telephony (D): detailed test results according to location. Adapted from [20] .... 63 
Table 19: Data (D): detailed test results according to location. Adapted from [20] ............ 65 

72
List of Abbreviations
2G Second generation
3G Third generation
3GPP 3rd Generation Partnership Project
AAA Authentication, Authorization and Accounting
ACR Anonymous Call Rejection
AMBR Aggregate Maximum Bit Rate
AMR Adaptive Multi-Rate
AMR-WB Adaptive Multi-Rate Wideband
API Application Programming Interface
APN Access Point Name
ARP Allocation and Retention Priority
ARPU Average Revenue Per User
AS Application Server
A-SBC Access Session Border Controller
AUTN Authentication Token
AVP Attribute Value Pair
AWS Advanced Wireless Spectrum
BGCF Border Gateway Control Function
BICC Bearer Independent Call Control
BSC Base Station Controller
BSS Base Station System
CAMEL Customized Applications for Mobile network Enhanced Logic
CDIV Communication Diversion
CDMA Code Division Multiple Access
CN Core Network
CONF Conferencing
CS Circuit Switched
CS NAS Circuit Switched Non-Access Signaling
CSCF Call Session Control Function
CSFB Circuit Switched Fall Back
CW Call Waiting
DEA Diameter Edge Agent
DiffServ Differentiated Services
DL DownLink
DNS Domain Name System
DPI Deep Packet Inspection
DRA Diameter Relay Agent
DRX Discontinuous Reception
DTM Dual Transfer Mode
ECGI E-UTRAN Cell Global Identifier
EDGE Enhanced Data rates for GSM Evolution
eNode B Evolved Node B

73
EPC Evolved Packet Core
EPS Evolved Packet System
ERAB E-UTRAN Radio Access Bearer
ESM EPS Session Management
eSRVCC Enhanced Single Radio Voice Call Continuity
ETSI European Telecommunications Standards Institute
E-UTRAN Evolved UTRAN
E-UTRAN Evolved Universal Terrestrial Access Network
FDD Frequency Division Duplex
G3 Fax Group 3 Fax
GAA Generic Authentication Architecture
GA-CSR Generic Access - Circuit-Switched Resources
GAN Generic Access Network
GBA Generic Bootstrapping Architecture
GBR Guaranteed Bit Rate
GERAN GSM EDGE Radio Access Network
GGSN Gateway GPRS Support Node
GPRS General Packet Radio Service
GSM Global System for Mobile Communications
GSMA GSM Association
GTP GPRS Tunneling Protocol
HLR Home Location Register
HPMN Home Public Mobile Network
HSPA High Speed Packet Access
HSPA+ Evolved High Speed Packet Access
HSS Home Subscriber Server
HTTP Hypertext Transfer Protocol
IBCF Interconnection Border Control Function
ICS IMS Centralized Services
I-CSCF Interrogating Call Session Control Function
ICSI IMS Communication Service Identifier
IETF Internet Engineering Task Force
iFC Initial Filter Criteria
IMEI International Mobile Equipment Identity
IM-GW IP Media Gateway
IMS IP Multimedia Subsystem
IMS-AGW IMS Access Gateway
IMS-AKA IMS Authentication and Key Agreement
IMS-ALG IMS Application Level Gateway
IMSI International Mobile Subscriber Identity
IOT Interoperability Testing
IP Internet Protocol
IP-CAN IP-Connectivity Access Network
IPsec IP Security

74
IP-SM-GW IP Short Message Gateway
IPX IP Packet Exchange
I-SBC Interconnect Session Border Controller
ISC IMS Service Continuity
ISIM IM Services Identity Module
ISUP ISDN User Part
IWF Interworking Function
LTE Long Term Evolution
MAC Medium Access Control
MBR Maximum Bit Rate
MCC Mobile Country Code
ME Mobile Equipment
MGCF Media Gateway Control Function
MGW Media Gateway
MIMO multiple-input multiple-output
MME Mobility Management Entity
MMS Multimedia Messaging Service
MMTel Multimedia Telephony
MNC Mobile Network Code
MNO Mobile Network Operator
MOS Mean Opinion Score
MRF Media Resource Function
MSC Mobile Switching Center
MSISDN Mobile Subscriber ISDN Number
MSRP Message Session Relay Protocol
MTU Maximum Transmission Unit
MWI Message Waiting Indicator
NAPTR Name Authority Pointer
NAS Non-Access Stratum
NAT Network Address Translation
NNI Network to Network Interface
OIP Originating Identification Presentation
OIR Originating Identification Restriction
OSS Operations Support System
OTT Over-The-Top
PCC Policy and Charging Control
PCEF Policy and Charging Enforcement Function
PCO Protocol Configuration Options
PCRF Policy Charging and Rules Function
P-CSCF Proxy Call Session Control Function
PDN Packet Data Network
PDN-GW PDN Gateway
PLMN Public Land Mobile Network
PS Packet Switched

75
PSHO PS handover
QCI QoS Class Identifier
QoS Quality of Service
RAB Radio Access Bearer
RAN Radio Access Network
RAND RANDom number (used for authentication)
RAT Radio Access Technology
RCS Rich Communication Suite
RES user RESponse (used in IMS-AKA)
RLC Radio Link Control
RNC Radio Network Controller
RNS Radio Network Subsystem
ROHC Robust Header Compression
RRC Radio Resource Control
RTCP RTP Control Protocol
RTP Real-time Transport Protocol
SAE System Architecture Evolution
SBC Session Border Controller
SCC AS Service Centralization and Continuity Application Server
S-CSCF Serving - Call Session Control Function
SCTP Stream Control Transmission Protocol
SDP Session Description Protocol
SEG Security Gateway
SGSN Serving GPRS Support Node
SGW Serving Gateway
SIGCOMP Signalling Compression
SIP Session Initiation Protocol
SMS Short Message Service
SPD Speech Path Delay
SRVCC Single Radio Voice Call Continuity
TAI Tracking Area ID
TAS Telephony Application Server
TCP Transmission Control Protocol
UE user equipment
UMA Unlicensed Mobile Access
UMTS Universal Mobile Telecommunications System
UTRAN UMTS Terrestrial Radio Access Network
VANC VoLGA Access Network Controller
VoIP Voice over IP
VoLGA Voice over LTE via Generic Access

76