The development of mobile is critically dependent on
`standards
`
`F Harrison and K A Holley
`
`The development of mobile standards has been a key factor in the widespread success and take-up of second generation (2G)
`cellular systems. Well thought through standards, including roaming aspects, have enabled operators, using equipment from
`different manufacturers, to create global mobile networks, which allow users to roam seamlessly and obtain largely the same
`services wherever they happen to be. Third generation (3G) standards have built on the success of 2G and have increased
`the global awareness and applicability. In this light, this paper discusses the background for mobile standards and the
`factors that are influencing their ongoing development for third generation and beyond.
`
`1. Mobile standards background
`
`The importance of standardisation to the mobile industry
`
`is probably best illustrated by the impact of the work on
`the second generation mobile systems, in particular the
`GSM (Global System for Mobile Communications)
`standard. The work was started in the mid-1980s, at about
`the same time as the launch of commercial first generation
`cellular networks. A variety of different first generation
`systems were deployed across Europe and the rest of the
`world. Network infrastructure and mobile terminals were
`relatively expensive and market adoption was mainly
`limited
`to business users. The objective of GSM
`standardisation was to create a new pan-European system,
`where users could continue to get service when roaming
`across country borders. Economies of scale of larger
`production volumes were expected to reduce prices.
`
`The impact of the GSM standards work is now clear to
`see. GSM is effectively a de facto world standard, claiming
`around 60% of the total mobile telephone population, and
`with coverage in all the continents. GSM terminals and
`network costs have reduced to such low levels that
`consumer pre-pay packages can be retailed at £50 or less,
`and the consumer market has developed such that the
`penetration of mobile telephones exceeds 50% of the
`population in many countries.
`
`The mobile communications industry has grown to the
`point where GSM manufacturers, such as Ericsson, Nokia
`and Motorola, and global operators, such as Vodafone, are
`now among the top companies in financial market
`capitalisation.
`
`is
`this development
`The speed and extent of
`undoubtedly due in large part to the creation of international
`standards. Had we continued by further developing the
`
`32
`
`BT Technol J Vol 19 No 1 January 2001
`
`differing national first generation standards, the mobile
`industry could not have come so far, so quickly.
`
`The challenge for third generation standards is to take
`this further — to create a truly single, world-wide standard.
`However, the standardisation environment has changed
`since the early work on GSM and it has been necessary to
`modify the approach. In particular, there are several key
`differences which have influenced the way in which 3G
`standards are being created:
`• GSM had been created by ETSI
`(European
`Telecommunications Standards
`Institute)
`as
`a
`European regional standard, and to create a world
`standard needs the involvement of other standards
`bodies, for example from the Far East and North
`America,
`the need to work on a world scale has introduced a
`wider variety of political and cultural influences on the
`standards work, where conflicting agendas can put the
`goal of a single standard at risk,
`substantial investments have now been made in second
`generation systems and infrastructure — whereas in the
`transition from first to second generation it had been
`possible to start with a clean sheet of paper, for third
`generation there is a strong commercial drive to reuse
`existing infrastructure and create an evolutionary
`approach, in order to capitalise on past investments,
`the extent of roaming capabilities with GSM is
`significant across the globe, and different networks
`will proceed at different rates towards third generation,
`but users of 3G will not accept a reduction in roaming
`capability and this requires intelligent handling of
`roaming between 3G and 2G — this situation will also
`lead to more complex combined 2G and 3G handsets,
`
`•
`
`•
`
`•
`
` Ex. 1107
`APPLE INC. / Page 1 of 6
`
`

`

`THE DEVELOPMENT OF MOBILE IS CRITICALLY DEPENDENT ON STANDARDS
`
`•
`
`the technology scope of 3G has broadened con-
`siderably to embrace data and Internet aspects — while
`GSM is a largely self-contained standard, for 3G it is
`necessary to draw on component standards from other
`bodies, such as the Internet Engineering Task Force
`(IETF) IP protocols, and this requires co-ordination
`and collaboration across different standards forums,
`where quite different standardisation processes are
`used.
`
`These and other aspects led, in December 1998, to the
`creation of a new partnership project known as 3GPP (3rd
`Generation Partnership Project), involving several regional
`standards bodies working together.
`
`2. Why standards are important
`
`The development of standards benefits users and all
`
`sectors of the mobile industry — suppliers, operators
`and regulators. Examples of these are given in Table 1.
`
`On the counter side, it is important that standards do not
`stifle innovation, and that product and service differ-
`entiation can still be achieved.
`
`In practice, such is the importance of standardisation to
`the mobile industry, that the creation of standards is a key
`part of the development process. New terminals and
`network systems simply cannot be launched without
`standards in place. One measure of the importance could be
`the scale of effort on 3G standardisation, where tens of
`thousands of engineers are already engaged in the standards
`process. At any time of any day, you could be reasonably
`certain that there is a mobile standards committee in session
`somewhere in the world.
`
`3. The 3G standard
`
`The 3G standard is being created against the general
`
`industry
`background of user expectations and
`imperatives. In theory, a single standard could perhaps be
`created with a single standards body, including all interested
`parties. In practice, there are different interests and starting
`points, in particular the desire to evolve from various
`second generation start points. Figure 1 shows the estimated
`market
`share
`(customer numbers, world-wide) by
`technology type. While the GSM technology dominates this
`chart, there are some important other starting points for
`evolution to 3G, including the TDMA (DAMPS) system,
`used widely in North America, PDC used in Japan and the
`CDMAone standard used in North America and the Far
`East. To take account of these interests, within the ITU the
`concept of a family of standards, known as IMT2000, was
`developed.
`
`Table 1Benefits of standardisation.
`
`Economies of scale
`
`Competitive procurement
`
`Service portability
`
`Roaming
`
`Regulatory
`
`Simplicity of service
`
`Service interworking
`
`Applications and services
`environment
`
`Spectrum efficiency
`
`Radio technology
`harmonisation
`
`Operators benefit from lower cost
`infrastructure.
`User benefit from low cost terminals.
`Suppliers benefit from large and stable
`market.
`Open standards enable interworking of
`network elements from different suppliers,
`avoiding long-term lock-in to a single
`supplier.
`Users can change to different terminal
`equipment or move between service
`provider and network operators while
`retaining a similar service.
`Services can be made available outside the
`user’s home network with the goal of
`having exactly the same look and feel.
`Regulators can specify standards for
`spectrum use, giving a controlled, yet open
`and competitive basis for licensing.
`Operators and suppliers can access new
`markets without being disadvantaged by
`local standards requirements.
`Services and terminals operate in a similar
`fashion, making it easier for users (e.g. use
`of + symbol for international calls).
`Standards can be used to ensure that
`services work well across different
`networks.
`Network and terminal standards enable the
`development of 3rd party applications, and
`allow the development of 3rd party value
`added services.
`Standards enable radio technologies to
`work together, or in close proximity within
`the radio spectrum, without undue
`interference.
`Standards are needed to ensure that
`different technologies can work in
`harmony in the same or similar coverage
`area without undue interference.
`
`1%
`
`11%
`
`12%
`
`9%
`
`8%
`
`GSM
`
`PDC
`
`TDMA
`
`CDMA
`
`analogue
`
`other
`
`59%
`
`Fig 1Estimate of market share by technology type.
`
`In fact a total of five different air interface standards are
`included in this IMT2000 family, these being:
`
`•
`
`CDMA-DS (direct sequence) — specified by 3GPP,
`also known as UTRA, FDD mode,
`
`33
`
`BT Technol J Vol 19 No 1 January 2001
`
` Ex. 1107
`APPLE INC. / Page 2 of 6
`
`

`

`THE DEVELOPMENT OF MOBILE IS CRITICALLY DEPENDENT ON STANDARDS
`
`•
`
`•
`
`•
`
`•
`
`CDMA-MC (multi-carrier) — specified by 3GPP2,
`also known as cdma2000,
`
`CDMA-TDD (time division duplex) — specified by
`3GPP, being a combination of the UTRA TDD mode
`together with the Chinese TD-SCDMA,
`
`TDMA single carrier — specified by 3GPP, also
`known as EDGE,
`
`FDMA/TDMA — specified by ETSI and based on the
`existing DECT standard.
`
`In addition, there are two different core network
`standards, one based on evolution from GSM and the other
`based on evolution from the North American IS41 standard.
`
`While this may seem like a perplexing range of options,
`there has been some effort to harmonise the different
`specifications, such that there is some alignment of the key
`technical parameters within the three CDMA modes, and
`there are developments to enable each air interface to be
`used with either of the core network standards. With this set
`of specifications, each major interest group can find an
`appropriate evolution path (see Table 2).
`
`Table 2Evolution options for different interest groups.
`
`DAMPS
`
`ATT Wireless
`
`2G investment Example operator* Evolution options
`GSM
`BT Wireless
`UTRA FDD and/or UTRA TDD
`with evolved GSM core.
`EDGE with GSM core.
`EDGE with evolved GSM or IS41
`core.
`cdma2000, with IS41 core.
`UTRA FDD with GSM evolved
`core.
`*Note — the ‘example operator’ column shows examples of operators
`with the respective 2G investment and does not imply any declared intent
`to follow any particular evolution path.
`
`CDMAone
`PDC
`
`Sprint PCS
`NTT DoCoMo
`
`From an operator perspective, while this approach gives
`a reasonable upgrade path to 3G, the variety of standards
`does present some problems. For global roaming, it may be
`necessary to support several standards, in order to allow
`access to service in those parts of the world where the home
`standard has not been implemented. This will require the
`use of multi-standard terminals, which currently represent a
`technology challenge, and are likely to have a cost
`premium. In addition, it will be necessary to develop cross-
`standard network interfaces and commercial arrangements.
`Despite this, the 3G standards landscape is a considerable
`improvement on
`the existing 2G multi-standard en-
`vironment.
`
`From a supplier perspective, the stabilisation of a
`standards family provides some good opportunities to enter
`into markets which have been difficult to access. For
`example, European suppliers have not been able to
`
`34
`
`BT Technol J Vol 19 No 1 January 2001
`
`penetrate into the Japanese market due to the unique
`standard in use in Japan, while Japanese suppliers have not
`capitalised on the opportunity to supply to the world GSM
`market, due to focus on their home market. However, the
`range of standards and standards bodies does present a
`significant resourcing challenge, for even the biggest
`suppliers. This is exacerbated by the fact that all standards
`options are developing in parallel.
`
`For regulators and governments, the openness of a 3G
`standards set helps in removing artificial trade barriers, and
`gives an opportunity to attract competitive overseas bidders
`to enter into the 3G licensing processes.
`
`4.
`
`3G standards developments — who is doing what?
`
`4.1
`
` 3GPP
`
`The GSM community, including ETSI and ANSI-T1 in
`
`the USA, along with several Far Eastern standards
`bodies (ARIB and TTC from Japan, TTA from Korea and
`CWTS from China), has been brought together by 3GPP
`[1]. The broad objective of 3GPP is the development of a
`3rd generation standard based on a direct sequence wide-
`band CDMA or time division duplex radio interface,
`coupled into an evolved core network based on GSM
`Release 98. This makes a significant advance on GSM by
`providing much higher data rates, but retains the GSM core
`network, allowing 2G/3G seamless roaming. Several initiat-
`ives which were already in place for GSM are also extended
`significantly, for example the SIM application tool-kit and
`the mobile execution environment (MExE), both of which
`allow a highly customisable terminal environment.
`
`4.2
`
`3GPP2
`
`3GPP2 [2] is another global initiative with support from
`TR.45 in the USA, ARIB and TTC from Japan, and TTA
`from Korea. It defines the cdma2000 standard which builds
`on the 2G CDMAone standard. Although data rates are
`increased roughly in line with the 3GPP standards, the core
`network is, however, still based on the IS41 core network,
`making roaming between cdma2000 and 3GPP systems
`difficult.
`
`4.3
`
`
`
` ITU
`
`The ITU [3] continues to have a role to play in the
`development of mobile standards. For some time, the ITU
`has had a programme of development work under the ‘IMT
`2000’ banner. When the 3GPP and 3GPP2 bodies emerged
`as strong developers of 3G standards, the ITU agreed to
`provide an umbrella framework so that these two standards
`bodies could be recognised by the ITU as developing the
`detailed specifications, along with ETSI for the enhanced
`DECT specification. Work inside the ITU and in groups of
`ITU members has resulted in significant harmonisation
`
` Ex. 1107
`APPLE INC. / Page 3 of 6
`
`

`

`THE DEVELOPMENT OF MOBILE IS CRITICALLY DEPENDENT ON STANDARDS
`
`efforts which have brought 3GPP and 3GPP2 radio aspects
`much closer together.
`
`4.43G.IP
`
`3G.IP was born in May 1999 out of a desire to bring
`mobile services firmly into the ‘Internet age’. Its primary
`goal was to put the success of 2G mobile systems,
`particularly those based on the GSM core network, together
`with the rapid growth of the Internet and the increasing
`interest in telephony and other real-time services over the
`Internet. It started modestly, with the objective to make
`rapid progress by bringing together like-minded companies
`(3G systems manufacturers and operators) who could
`develop the core principles for transporting IP-based
`multimedia services over a 3GPP system based on EDGE or
`UTRA.
`
`3G.IP was very successful in establishing an agenda
`towards ‘all-IP’ and helping 3GPP to develop the first draft
`architecture, the first step towards a complete IP multimedia
`system. Once documents with proposals started to find their
`way into 3GPP, however, it became clear that 3G.IP could
`not make the desired progress without becoming much
`more open and allowing other companies to participate. So
`3G.IP started to admit other companies towards the end of
`1999, becoming completely open early in the year 2000.
`3G.IP has no subscription charges and the only prerequisite
`for joining is the signing of an agreement which may be
`downloaded from its Web site [4].
`
`4.5IETF
`
`The IETF [5] is focused on the development of
`protocols which run over an IP stack. As mobile standards
`migrate more and more towards IP-based services it
`becomes more important that the IP service standards take
`mobile into consideration. There is currently much interest
`in the IETF in further development of IETF protocols
`relating to mobile, including call-control protocols such as
`SIP, mobility services such as mobile IP, and transport
`efficiency aspects such as IP header compression. The IETF
`does not consider system aspects and therefore a close
`dialogue between the mobile standards community and the
`IETF is required to produce the right standards which help
`the system to work effectively.
`
`4.6WAP Forum
`
`In 1997 several different manufacturers started to
`develop proprietary methods for converting Internet content
`into a more compact format for transmission over expensive
`and bandwidth-limited mobile data channels. It was during
`a session in a subgroup of ETSI SMG, responsible for the
`GSM standards at that time, that some of these manu-
`facturers presented their ideas to the standards community.
`The ETSI standards community was
`interested, but
`
`disappointed that these ideas were not brought into ETSI, as
`it was obvious that operators could not support multiple
`different protocols according to the handsets being used at
`the time by their subscribers. So these manufacturers were
`given a clear message that they must work together or their
`ideas would not take off. The WAP Forum [6] was born
`later that year, with the objective of defining a standard
`‘micro-browser’ to run on mobile handsets, and a suite of
`protocols which allow that micro-browser to interact with a
`network server. WAP micro-browsers can be implemented
`on a variety of 2G handsets, including GSM, CDMAone
`and US TDMA handsets. During the year 2000, WAP
`telephones have started to appear on the market and the
`WAP Forum is moving on to providing a richer en-
`vironment for delivering Internet content to mobiles.
`
`4.7Mobile Wireless Internet Forum (MWIF)
`
`While 3G.IP focused on the delivery of an IP
`multimedia system which builds on 3GPP Release 99,
`MWIF [7] started early in the year 2000 to study how the
`existing Internet paradigm could be made mobile. MWIF
`members come from 3G operators and manufacturers,
`including mainstream Internet systems suppliers. MWIF
`also includes the ISP community. MWIF is now producing
`a proposal for a functional architecture which will be fed
`into 3GPP and 3GPP2 later this year. The MWIF approach
`is to define a solution which is independent of access
`technology and is, in general, much more radical than has
`been discussed in 3GPP to date, hinting at changing as
`many as possible of the protocols from today’s tele-
`communications protocols to IP-based protocols.
`
`5. The move to IP architecture and standards impact
`
`Over the past couple of years there has been increasing
`
`interest in moving to generalised IP-based architectures
`and protocols and avoiding expensive specialist protocols
`and architectures for different implementations of the same
`service for mobile, fixed, tetherless, etc. In the 3G standards
`world, one of the most significant impacts is the use of
`‘voice over IP’ technology within the mobile networks.
`3G.IP developed an initial architecture which put together
`the IP call server approach used by both H.323 and SIP with
`the mobile architecture developed for 3GPP. Instead of
`carrying voice traffic through expensive MSCs, IP links are
`established to call servers using the basic IP bearer
`capability of the mobile network. The call server can then
`perform largely the same job as in the ‘fixed’ ISP-based
`networks, instructing a gateway to set up a call using ‘voice
`over IP’ or over the PSTN, and then instructing the mobile
`to establish another IP bearer channel with higher
`bandwidth to the gateway for transport of voice traffic.
`
`For 3GPP, however, it is not as simple a process as this.
`There are a number of aspects which need detailed
`consideration inside the 3GPP standards process:
`
`35
`
`BT Technol J Vol 19 No 1 January 2001
`
` Ex. 1107
`APPLE INC. / Page 4 of 6
`
`

`

`THE DEVELOPMENT OF MOBILE IS CRITICALLY DEPENDENT ON STANDARDS
`
`•
`
`•
`
`•
`
`the security and authentication mechanisms have to be
`examined to ensure that fraud and eavesdropping are
`avoided, while maintaining the ability to invoke lawful
`interception,
`
`the IP bearer channels for both signalling and voice
`traffic need to be of high enough quality to ensure that
`calls proceed without failure and that voice quality is
`equivalent to the circuit-switched world,
`
`the service or feature interaction between the call
`control model and additional services provided by
`operators, such as prepay, VPN, number translation,
`etc, needs to be handled to offer the same look and feel
`as the circuit-switched world,
`
`• when roaming, all interactions must be handled
`smoothly so that the user does not perceive any
`difference between
`roaming on circuit-switched
`networks and roaming on IP-based networks.
`
`The overall architecture for 3GPP has now been
`defined, and the additional functions required in the 3GPP
`core network have been termed the ‘IP multimedia sub-
`system’.
`
`Having addressed these considerations, there are the
`specific protocols to be considered. 3G.IP made an early
`decision to press for the use of the IETF session initiation
`protocol (SIP) as a standardised protocol to be used between
`the user equipment and the call server. This decision was
`also adopted by 3GPP. In principle, this means that the
`IETF standard SIP can just be slotted into place, but, in
`practice, there will need to be changes made to SIP to
`accommodate the needs of a mobile network. 3GPP has
`decided that it will encourage the development of these
`extensions within the IETF rather than making a 3GPP-
`specific version of SIP. Since the IETF will concern itself
`only with the protocol aspects, it is necessary for 3GPP to
`develop the overall system framework and develop the
`necessary additional function descriptions before the IETF
`can work out the details of the required extensions.
`
`Overall, this means that the mobile standards landscape
`is moving from a position where all mobile standards for a
`particular system are developed in one standards committee
`structure to one where different standards structures must
`co-operate to develop the complete solution. The IP
`architecture is a particular challenge because the companies
`involved have to bring together small low-power mobile
`devices having hard-coded implementations with fast-
`moving IP standards which expect the latest code to be
`downloadable on to the end-user device.
`
`36
`
`BT Technol J Vol 19 No 1 January 2001
`
`6. Conclusions and future outlook on mobile
`standards
`
`There can be little doubt that the pace of change in
`
`mobile is accelerating. Can the standards process keep
`up and continue to deliver the benefits of standardisation
`without delaying the introduction of new products and
`services? The future success of standardisation will be
`dependent on the ability to position standards in order to
`protect investments while giving freedom to innovate with
`new developments.
`
`There is an increased speed of change and an
`expectation of ever shorter development lead times.
`However, this needs to be counterbalanced by the very high
`investment costs in introducing new network technologies,
`e.g. in the UK alone, the 3G licence cost to BT Wireless of
`£4bn, plus an expected network investment of a further
`£2bn, mitigates against a high rate of change. Putting
`together the world-wide costs of 3G and taking into account
`all of the different competing operators across all regions
`will result in a cost running to many hundreds of billions of
`pounds.
`
`The key is to provide basic standards which provide a
`sound and stable platform, protecting the fundamental
`investments, while enabling a change of functionality and
`capabilities quickly and at low incremental costs. Some of
`the current developments will help us move to this
`paradigm:
`
`•
`
`•
`
`•
`
`creation of linked standards families, giving evolution
`options and choices for implementation, e.g. the reuse
`of the GPRS core network in the 3GPP standards suite
`enables service and investment continuity during the
`transition to 3G — for this to work well, a long-term
`vision is required, such that the basic architecture
`decisions support future evolution,
`
`the move to IP-based architectures, where services
`applications can be developed quickly and supported
`on a simple, universal protocol structure — the
`creation of application programming interfaces (APIs)
`at both the network and the terminal is also a key
`enabler for this,
`
`the emergence of techniques such as software radio,
`where
`terminals and
`infrastructure can be
`re-
`configured by downloading or software upgrade to a
`new standard.
`
`Another factor relating to the rate of change is the way
`in which standards are developed. This is now a resource-
`intensive (but vital) part of the development process.
`Already, standards are created
`in cyberspace, with
`electronic meetings, e-document development and e-voting
`processes. Even so, the lead time for significant standards
`developments, such as CAMEL, GPRS or 3G is still of the
`
` Ex. 1107
`APPLE INC. / Page 5 of 6
`
`

`

`THE DEVELOPMENT OF MOBILE IS CRITICALLY DEPENDENT ON STANDARDS
`
`order of 3—5 years. Even relatively small enhancements
`can take a year to standardise and a further year before
`deployment. So, one challenge for the industry is to
`streamline and automate the standards process, with
`increased use of development tools.
`
`3 ITU Web site — http://www.itu.int/
`
`4 3G.IP Web site — http://www.3gip.org
`
`5 IETF Web site — http://www.ietf.org
`
`Standards can no longer be developed by small isolated
`committees. To produce high-quality mobile standards
`resulting in high-quality global mobile systems with full IP
`multimedia capability requires development within a
`number of different global standards organisations. The
`management and co-ordination of input into all the various
`forums is a big challenge for all companies in the industry,
`and, in particular, ensuring that deals struck between
`companies in one forum are upheld in a different forum will
`be very testing for all concerned. It is only the major
`standards players who will be in a position to drive results
`across all communities and create the right global mobile
`standards to create and enhance value for their shareholders.
`
`Can standards keep up, and be created fast enough to
`meet commercial needs? Well, it is the commercial needs
`which drive the speed of standardisation, rather than vice-
`versa. The difference perhaps is that standards are driven by
`the overall industry commercial needs, often being a
`compromise, and perhaps delaying those who would benefit
`from a very high standards churn rate.
`
`Overall, the outlook for mobile standards is highly
`positive — they are an intrinsic and essential component of
`network and terminal developments, and can clearly be seen
`to have played a major part in the success and growth of the
`mobile industry world-wide.
`
`References
`
`1 3GPP Web site — http://www.3gpp.org
`
`2 3GPP2 Web site — http://www.3gpp2.org
`
`6 WAP Forum Web site — http://www.wapforum.org
`
`7 MWIF Web site — http://www.mwif.org
`
`Fred Harrison works in BT Wireless, where
`he is responsible for directing work on 3rd
`generation mobile standards. In 1998/99 he
`took a leading role in establishing the new
`third generation partnership project standards
`body (3GPP). He has also been active in the
`creation of the 3G.IP Focus Group.
`
`He started his career with BT in 1974 as a
`sponsored student. He studied Electronic
`Engineering at Southampton University,
`graduating in 1978. He worked on optical
`fibre and microwave radio developments
`until 1990 when he moved to Cellnet, where
`he was responsible for the overall design and planning of Cellnet’s radio
`network. He then spent two years in BT Group Technology prior to the
`formation of BT Wireless.
`
`from Bristol
`Kevin Holley graduated
`University in 1985 with a BSc in Physics. He
`joined BT Laboratories that year and he
`began working on off-air analysis of analogue
`cellular signalling. His involvement in the
`global system for mobile communications
`(GSM) standards work began in 1988 with a
`study of the reliability of the proposed
`signalling mechanism. Later in 1988 he
`began work on the short message service
`(SMS). In 1989 he was seconded to Cellnet to
`work on GSM technology development. He
`returned to BT at Adastral Park in 1992, but
`has continued his involvement in GSM
`standardisation and development right up to the present day third generation
`(3GPP) standards where he is mostly focused on the terminal aspects.
`
`37
`
`BT Technol J Vol 19 No 1 January 2001
`
` Ex. 1107
`APPLE INC. / Page 6 of 6
`
`