Alcatel-Lucent continues to shift towards pole position in the £10 billion ($20 billion) 21Century Network project at British Telecom, one of the world’s most advanced programs to migrate to an all-IP converged network.
The latest win for the French giant is a contract to support the automation of customer migration from the legacy network to 21CN, a process that is just beginning. However, the new deal is a small part of a far larger vendor battle to maximize its share of BT’s huge investment, one that will be replayed with many other large carriers around the world. And once the actual equipment dollars are allocated, the real prize will be the service delivery architectures, including near term SDPs (service delivery platforms) and the ultimate goal of an all-encompassing IMS (IP Multimedia Subsystem).
Although IMS has tended to be driven by the wireless carriers, this is starting to change as the incumbents get serious about all-IP deployment, and BT – like converged counterparts such as France Telecom and AT&T – is taking a leading role in defining what IP services delivery will look like in reality. This is a process that should accelerate the availability of usable IMS from the involved vendors, with Alcatel-Lucent one of the major frontrunners, but could also raises the specter of fragmentation, as each operator seeks to ensure competitive advantage by shaping its IMS in its own commercial image.
As well as supporting a general move to VoIP as the main voice engine, 21CN will support all BT’s planned next generation applications, across wired and – through spectrum acquisition and/or partnerships – wireless links.
Mobile television, location and presence awareness and enterprise multimedia services are in the front line for 21CN offerings, and like other carriers, BT is focusing heavily on new classes of devices that can support new ways of communicating and interacting, with fully usable mobile web interfaces.
One of its strongest signals of its strategy in this area was its recent alliance with Sony to bring the PSP handheld console into the 21CN ecosystem. This enables Sony to tap into some of the advanced R&D that BT has carried out into VoIP and video over IP in the process of developing its all-IP network. Such expertise will be important in turning products like the PSP into real IP workhorses, ushering in a new set of user behaviors, and a new relationship between device makers and operators.
The other key area of interest for all-IP carriers like BT is Web 2.0, with its potential to support a new breed of potentially revenue generating behaviors, and the UK carrier has been advanced in seeking to harness the emerging developer community for Web 2.0 and mobile web services at an early stage. It has pipped almost all its counterparts to the post in issuing a software developers’ kit (SDK) for 21CN, in a bid to start to develop and showcase potential services, and put some concrete behind the large claims for how all-IP networks will revolutionize the user experience – for instance, BT said in a statement last week that ‘soon bandwidth will no longer be a consideration for most end users as additional bandwidth can be requested as and when required’.
That is fine, but there need to be strong applications to use that bandwidth and the SDK should help to generate interest and innovation. France Telecom is also taking the view that applications development must not wait for the moment when the IMS and other architectures are finalized, but must be generated in parallel. Its three major focuses of mobile Web 2.0 software R&D are currently focused on user generated content and music, with a prototype service called Soundtribes; new methods of organizing vast stores of information on the web, with a project called Bubbletop; and ‘geo-tagging’ or associating photos and other data to maps, in which France Telecom has a product called Pikeo in its Orange Labs research nerve center.
Meanwhile, the operators are also preoccupied with the most efficient ways to support, provision and deliver these new services and applications. Many are looking to do this initially using currently available SDPs, such as Microsoft’s Connected Services Framework (CSF – in use at BT and AT&T), before migrating these systems to form part of a broader IMS around the turn of the decade, when many telcos, BT included, aim to turn off their PSTN networks altogether. Alcatel-Lucent is providing the IMS for 21CN.
Two problems are afflicting IMS as far as carriers are concerned, both related to the sheer size and scope of the fully blown standard. One is the risk associated with the kind of ‘big bang’ deployment that vendors like Ericsson dream of. This has led to a sharpened operator interest in technologies like SDP and UMA (Unlicensed Mobile Access), which can provide some of the functions of IMS immediately but with lower risk and with full integration with legacy systems, and can be incorporated into a future IMS strategy. This, in turn, has shifted the focus of some vendors, most notably Cisco, to providing step-by-step IMS architectures.
The other challenge is that the IMS, if implemented in vanilla standard form, falls into the trap of most complex standards – meeting most of the requirements of most of the operators, but failing to be perfect for any of them. This has led various large carriers to work with their favored vendors on creating a version of the platform specifically optimized for their own business and technological priorities.
Most of this work has taken place in the US, with Verizon’s A-IMS and AT&T’s Carts. Verizon Wireless is working with five key suppliers – Cisco, Qualcomm, Motorola, Nortel and Alcatel-Lucent – on an enhanced implementation of the CDMA version of IMS (called MMD), which is designed to fill gaps in existing 3GPP2 standards.
Verizon apparently finds this platform lacking in two ways: gaps in the specifications, caused by new applications requirements that were not anticipated until recently; and an insufficiently evolutionary path from current systems to the key IP services protocol, SIP. The A-IMS architecture will be the basis of next generation services but can also be implemented on top of current networks, a key factor for migration.
Meanwhile, AT&T is working on a document to be circulated to telecoms and IT suppliers, defining its own next generation services platform and going significantly further than pure IMS. According to an interview conducted with Light Reading, the document will set out a refined version of Carts (Common Architecture for Real Time Services), an AT&T specification that describes all the middleware that sits between the carrier’s physical fixed and mobile network elements and the customer applications.
Like IMS, this is designed to allow customers to choose services independently of access technology or device and to switch between access technologies seamlessly while in live sessions – for instance, to gain a stronger wireless signal or when a cheaper link comes into range. The main reason for updating Carts is to account for the integration of BellSouth and Cingular into AT&T.
AT&T (in its previous incarnation as a long distance carrier, pre-SBC takeover) originally developed the Carts approach as a practical manifestation of its ‘Concept of One’ idea, its strategy to move towards a single MPLS-based network to unify all its services. The telco then set about creating an architecture to support this, before IMS standards were sufficiently advanced to do the job – this included the RSOIP (Real Time Services over IP) technology, which looks very much like a prototype for IMS.
And like Verizon, AT&T is unwilling to let go of the enhancements its own inhouse work has brought, so instead has created Carts as the extension of RSOIP, and is incorporating standard IMS functions as a subset, primarily for session management. Carts also includes its own service delivery platforms (currently AT&T is using CSF) and the back end OSS processes.
In response to the first Carts document, two vendors have already delivered a key element, the Service Logic Execution Environment, according to Light Reading. This is based on SIP and HTTP and supports interaction between SIP and Web 2.0, for applications that involve complex web-based interactions.
The AT&T and Verizon initiatives will undoubtedly contribute major knowhow to the IMS effort – Verizon is looking to create an industry body and gain the support of other operators for its developments – but they also raise fears of fragmentation of IMS, and provide yet another question mark over whether the standards process is any longer adequate to the complex commercial challenges and time to market pressures of the converged world.
Last month, the operator-driven Open Mobile Terminal Platform group also set to work on a document containing specifications for IMS, and claiming to fill gaps in the standard and make it more deployable by real world carriers. Tim Raby, CEO of the OMTP, said that IMS will offer many benefits, especially improved quality of service, but that ‘for this to become a reality however there is a certain amount of ‘knitting’ needed to join up various elements. Our requirements give clear guidance to ensure IMS has the opportunity to deliver on its promises.’
The standards bodies are seeking to allay fears of the fragmentation that has held back other software architectures, such as some Java frameworks, in the mobile and converged arenas. Recently, 3GPP gained the support of European body ETSI for a common development of IMS Version 8, bringing together the 3GPP work and that of the ETSI TISPAN (Telecoms & Internet converged Services and Protocols for Advanced Networks) project, which had looked to emulate the mobile-focused standard for fixed line systems.
However, the end results still need to be sufficiently rich and usable to underpin the operators’ ambitious IP plans and the US carriers, in particular, remain to be convinced – even while their European equivalents, particularly BT and France Telecom, seem more prepared to stick with IMS standards and look to applications development and ecosystems to create their differentiation in the all-IP world.