Integration of UMTS and B-ISDN

is it possible or desirable?

Lloyd Wood,
student, MSc in Satellite Communication Engineering,
Department of Electronic and Electrical Engineering,
University of Surrey.

Assignment submitted as coursework to Ms F. Coakley, 17 March 1995.


Contents


Introduction

In the future, existing fixed networks will be complemented by mobile networks with similar numbers of users. These mobile users will have identical requirements and expectations to the fixed users, for on-demand applications of telecommunications requiring high bit-rate channels. It will be necessary for these fixed and mobile networks to interoperate in order to pass data, in real time and at high speeds, between their users.

But how far must this interoperation be taken? How much integration of the fixed and mobile network structures is needed? Here, a fixed network, B-ISDN, and a mobile network, UMTS, under development at the same time, are examined to see how well and closely they should work together in order to meet expected user needs. Work already taking place on this is discussed.

Background

The Universal Mobile Telecommunication System (UMTS), the third generation of mobile networks, is presently being specified as part of the European RACE technology initiative. The aim of UMTS is to implement terminal mobility and personal mobility within its systems, providing a single world mobile standard. Outside Europe, UMTS is now known as International Mobile Telecommunications 2000 (IMT2000), which replaces its previous name of Future Public Land Mobile Telecommunication System (FPLMTS). [BUIT95]

UMTS is envisaged as providing the infrastructure needed to support a wide range of multimedia digital services, or teleservices [CHEU94], requiring channel bit-rates of less than the UMTS upper ceiling of 2 Mbits/second, as allocated to it in the World Administrative Radio Conference (WARC) '92 bands. UMTS must also support the traditional mobile services presently offered by separate networks, including cordless, cellular, paging, wireless local loop, and satellite services. [BUIT95] Mobile teleservices requiring higher bit rates, from 2 to 155 Mbits/second, are expected to be catered for by Mobile Broadband Services (MBS), the eventual successor to UMTS, which is still under study. [RACED732]

Broadband Integrated Services Digital Network (B-ISDN), conceived as an all-purpose digital network that will supersede Narrowband ISDN (N-ISDN or ISDN), is also still being specified. B-ISDN, with its transport layer of Asynchronous Transfer Mode (ATM) is expected to be the backbone of future fixed digital networks. [MINZ89]

It is anticipated that, by the year 2005, up to 50% of all communication terminals will be mobile. [CHEU94] The Mobile Green Paper, issued by the European Commission in 1994, predicts 40 million mobile users in the European Union by 2000, rising to 80 million by 2010. This gives mobile users an importance ranking alongside fixed-network users. [BUIT95]

One result of this growth in mobile telecommunications will be the increase in teleservice operations that originate in either the fixed or mobile network, but terminate in the other, crossing the boundary between the two. UMTS is expected to be introduced within the next ten years, and integration with narrowband and broadband ISDN is possible in this time. Interoperability between UMTS and ISDN in some fashion will be necessary to support the interoperability between the fixed and mobile networks that users have already come to expect with existing mobile networks, and to meet the expectation of consistency of fixed/mobile service provision laid out in the initial RACE vision. [SWAI94]

One way of making UMTS attractive to potential customers is to offer the same range of services that B-ISDN will offer, within the bounds of the lower 2 Mbits/second ceiling of UMTS. [BUIT95]

So, with the twin goals of meeting existing expectations and making UMTS as flexible as possible to attract customers, how closely integrated must UMTS be with B-ISDN to achieve this?

Alternatives for integrating UMTS with other networks

The UMTS network could be developed along one of the following alternative integration paths:
  1. Developing an 'optimised' network structure and signalling protocols tailored for the special mobile requirements of UMTS. This would be incompatible with anything else. Services from all fixed networks would be passed through via gateways. This design-from-scratch method would result in highly efficient intra-network operation, at the expense of highly inefficient inter-network operation, high development cost, scepticism relating to non-standard technology, and slow market take-up. True integration with fixed networks is not possible in this scenario.

    Given the drawbacks, this is not a realistic option, and it has not been considered in depth. One of the RACE goals was to design UMTS not as a separate overlay network, but to allow integration with a fixed network; this option is undesirable. [BUIT95]

  2. Integration with and evolution from the existing Global System for Mobile telecommunication. (GSM, formerly standing for Group Special Mobil during early French-led specification, is now taken as meaning Global System for Mobile communications by the non-French-speaking world.) GSM is currently being introduced on the European market.

    This option has the advantage of using already-existing mobile infrastructure with a ready and captive market, but at the expense of limiting channel bit-rate considerably, which in turn limits the services that can be made available over UMTS. Some of the technical assumptions of UMTS, such as advanced security algorithms and distributed databases, would require new protocols to implement over GSM. GSM would be limiting the capabilities of UMTS. [BROE93a]

  3. Integration with N-ISDN. Like the GSM option above, this initially limits UMTS's channel bit-rate for services, but has a distinct advantage over integration with B-ISDN - N-ISDN is widely available, right now. However, integrating UMTS and N-ISDN would require effective use of the intelligent network concept for the implementation of mobile functions, and modification to existing fixed network protocols to support mobile access.

    Integrating UMTS with N-ISDN makes possible widespread early introduction and interoperability of UMTS in areas that do not yet have B-ISDN available. This allows wider market penetration, as investment in new B-ISDN equipment is not required, and removes the dependency of UMTS on successful uptake of B-ISDN for interoperability with fixed networks. Eventual interoperability with B-ISDN, albeit with constrictions imposed on UMTS by the initial N-ISDN compatibility, is not prevented. [BROE93a]

  4. Integration with B-ISDN. This scenario was the target of MONET (MObile NETwork), or RACE Project R2066. Unlike the above options, B-ISDN's high available bandwidth and feature set does not impose limitations on the service provisioning in UMTS. Fewer restrictions are placed on the possible uses and marketability of UMTS as a result. Development of B-ISDN is taking place at the same time as UMTS, making smooth integration and adaptation of the standards to each other possible.

For these reasons, integration of UMTS with B-ISDN has been accepted as the eventual goal for interoperability of future fixed and mobile networks using these standards, and this integration has been discussed in depth. [BROE93a, BROE93b, BUIT95, NORP94]

At present, existing B-ISDN standards cannot support the mobile-specific functions required by a mobile system like UMTS. Enhancements supporting mobile functions, such as call handover between cells, are needed before B-ISDN can act as the core network of UMTS.

Flexible support of fixed, multi-party calls, to allow B-ISDN to be used in conferencing and broadcasting applications, has many of the same requirements as support for mobile switching, so providing common solutions to allow both could minimise the number of mobile-specific extensions that B-ISDN needs.

As an example of how B-ISDN can be adjusted to meet UMTS's needs, let's look at that mobile requirement for support for call handover. Within RACE a multiparty-capable enhancement of B-ISDN, upwardly compatible with Q.2931, has already been developed, and implementing UMTS with this has been studied. For example, a UMTS handover can be handled as a multi-party call, where the cell the mobile is moving to is added to the call as a new party, and the old cell is dropped as a party leaving the call, using ADD(_party) and DROP (_party) primitives. Other mobile functions can be handled by similar adaptations to the B-ISDN protocols.

The enhancements to B-ISDN Release 2 and 3 that are required for UMTS support are minimal enough to be able to form an integral part of future B-ISDN standards, without impacting on existing B-ISDN work. [BUIT95]

These modifications only concern high-level B-ISDN signalling protocols, and do not alter the transport mechanisms. The underlying ATM layers, including the ATM adaptation layer (AAL) are unaffected by this.

The intelligent network

The Intelligent Network (IN) is a means for service providers to create new services and rapidly introduce them on existing networks. As the IN was considered useful for implementing mobility procedures in UMTS, it was studied as part of MONET, and is now specified in the Q.1200 series of the ITU-T recommendations.

The intelligent network separates service control and service data from basic call control. Service control is then activated by 'trigger points' in the basic call. This means that services can be developed on computers independent of the network switches responsible for basic call and connection control. This gives flexibility to the network operators and service providers, as well as the potential to support the services on any network that supports the trigger points. Eventually, IN can be expanded to control the network itself, such as handling all UMTS mobile functions. [BROE93a]

Any network supporting the intelligent network service set will be able to support new services using that service set easily, making integration of networks easier and transparent to the user of those services. The intelligent network is thus an important factor in the integration of B-ISDN and UMTS. UMTS, B-ISDN and the intelligent network set are all being developed at the same time, allowing each to influence the others in producing a coherent, integrated whole. [BUIT95]

Conclusion

In order to be accepted by users as useful and to provide as wide a variety of services as possible, UMTS needs some form of interoperabilty or integration with a fixed network. Integration of UMTS with B-ISDN offers the most flexibility in providing services when compared to other network integration options, and constrains UMTS the least.

With the increase in the number of services that will be made available in UMTS and B-ISDN over present standalone services, it is unrealistic to develop two separate, and incompatible, versions of each service for the fixed and mobile networks. Integrating UMTS and B-ISDN makes the same service set available to both sets of users in the same timescale, reducing development costs for the services, and promoting uptake and use in the market. The intelligent network concept allows the easy provision of additional services with little extra development cost. Integrating UMTS with B-ISDN, and with the intelligent network set, is therefore desirable.

Work on this integration indicates that the mobile requirements of UMTS can be met by extending existing B-ISDN signalling to handle them, without significantly modifying B-ISDN. Integration of UMTS with B-ISDN is therefore technically feasible.

References

[BROE93a]
W. van den Broek, A. N. Brydon, J. M. Cullen, S. Kukkonen, A. Lensink, P. C. Mason, A. Tuoriniemi,
"RACE 2066: Functional models of UMTS and integration into future networks",
IEE Electronics and Communication Engineering Journal, June 1993.

[BROE93b]
W. van den Broek and A. Lensink,
"A UMTS architecture based on IN and B-ISDN developments",
Proceedings of the Mobile and Personal Communications Conference, 13-15 December 1993.
IEE Conference Publication 387.

[BUIT95]
E. Buitenwerf, G. Colombo, H. Mitts, P. Wright,
"UMTS: Fixed network issues and design options",
IEEE Personal Communications, February 1995.

[CHEU94]
J. C. S. Cheung, M. A. Beach and J. P. McGeehan,
"Network planning for third-generation mobile radio systems",
IEEE Communications Magazine, November 1994.

[MINZ89]
S. E. Minzer,
"Broadband ISDN and Asynchronous Transfer Mode (ATM)",
IEEE Communications Magazine, September 1989.

[NORP94]
T. Norp and A. J. M. Roovers,
"UMTS integrated with B-ISDN",
IEEE Communications Magazine, November 1994.

[RACED732]
IBC Common Functional Specification, Issue D.
Race D732: Service Aspects.

[SWAI94]
R. S. Swain,
"UMTS - a 21st century system: a RACE mobile project line assembly vision"

Following work


Lloyd Wood (L.Wood@society.surrey.ac.uk)
last updated 27 July 1999