IMS Guide

IMS is a label that has started springing up all over the telecom industry, whether on individual products, or as vendor systems platforms or interoperability programs. It’s the acronym for IP Multimedia Subsystem, and it is very, very important – a big vision of the future of telecommunications hangs on it. It’s also very, very complicated and pretty abstract.

This tutorial aims to resolve some of that unfortunate conjunction by providing a quick and easy guide to some questions that are frequently asked about IMS. In the report's initial form, answers to a number of basic questions are given, one per page. But the idea is that readers can ask further questions on the message board attached to this article. If you want to send a private message, please email us, and include "IMS Guide" in the subject field. Frequently posed questions will be answered by adding pages to this report.

For a basic understanding of IMS, here are the key starting points:

Support for IMS even at this early stage in its development seems to be fairly hard-nosed, if the results of a poll taken during the Light Reading Webinar on which some of this report is based – IMS: A Blueprint for Fixed-Mobile Convergence – are typical:

In other words, a large majority in favor of pragmatic flexibility, rather than any fancy and/or more futuristic possibilities.

Here's a hyperlinked list of the questions and answers in this report:

This report is based partly on material drawn from various Light Reading resources, including:

Archived Webinar on IMS:

Paid research reports on IMS:

Light Reading news analysis on IMS:

To learn more about IMS, check out the coming Light Reading Live! conference:

IMS: Blueprint for Fixed-Mobile Convergence
at The Sheraton Hotel (Buckhead) in Atlanta, on Wednesday, March 30, 2005

— Tim Hills is a freelance telecommunications writer and journalist

IMS stands for IP Multimedia Subsystem, a concept developed and specified by the 3rd Generation Partnership Project (3GPP). 3GPP is a collaboration agreement established in December 1998 among various standards bodies, currently (March 2005) comprising:

The original aim of 3GPP was to specify a third-generation mobile system based on evolved GSM core networks and the radio access technologies they support, but this was subsequently extended to include the maintenance and development of the GSM technical specifications and technical reports, including evolved radio access technologies. IMS is a result of this program.

IMS increases the functionality of packet-switched mobile networks (such as 3G GSM) by supporting IP-based applications and services through the SIP protocol. However, the rapid spread of fixed-network broadband and the offering of services such as transactions, content distribution, and VOIP over all-IP networks have made IMS increasingly relevant to fixed operators as well.

Effectively, IMS provides a unified architecture that supports a wide range of IP-based services over both packet- and circuit-switched networks, employing a range of different wireless and fixed access technologies. A user could, for example, pay for and download a video clip to a chosen mobile or fixed device and subsequently use some of this material to create a multimedia message for delivery to friends on many different networks. A single IMS presence-and-availability engine could track a user’s presence and availability across mobile, fixed, and broadband networks, or a user could maintain a single integrated contact list for all types of communications.

A key point of IMS is that it is intended as an open-systems architecture: Services are created and delivered by a wide range of highly distributed systems (real-time and non-real-time, possibly owned by different parties) cooperating with each other. It is a different approach to the more traditional telco architecture of a set of specific network elements implemented as a single telco-controlled infrastructure.

IMS is potentially the base of a new telecom business model for both fixed and mobile networks and is a key enabler of fixed/mobile convergence. In principle it replaces the traditional walled-garden approach of a single operator, offering a limited range of services from within a highly controlled network, with an almost limitless range of highly functional services that span multiple operator and service-provider domains – fixed and mobile.

Obviously, technical potential and industry politics frequently collide, and IMS is not necessarily going to create an open-service Nirvana overnight. During the standard’s development there was a strong desire to perpetuate the mobile walled-garden approach, while allowing third-party developers to contribute applications and application segments. 3GPP was very anxious for mobile service providers not to end up as bit-pipe suppliers.

Nevertheless, it’s the fixed and mobile aspect that has got many in the industry so excited. A virtuous circle has emerged in which developments that began in the wireline environment with IP and SIP were picked up by 3G standards developers, refined and extended, and now are being fed back to the wireline community. IMS may have been designed for 3G, but now it’s much more than that, and everyone is converging on the same basic platform: IMS.

Key attractions of IMS are:

IMS belongs to an ongoing suite of 3GPP standards for 3G mobile networks that began with Release 99 in 2000 and saw Release 4 in 2001 and Release 5 in 2004. IMS is specified for implementation in Release 5, and the basic architecture is described in 3GPP TS 23.002, which introduces the key CSCF concept. However, Release 5 is not yet deployed in 3G networks, and IMS-enabled applications, such as POC, are being deployed in pre-R5 networks based, for example, on R99.

The standards continue to evolve, and the architecture reached Version 6.6.0 in December 2004. This version is not approved by the 3GPP, and so is not for implementation; rather, it is to aid future development of IMS within 3GPP. Release 6, when finally approved, will significantly extend IMS capabilities, making it easier to deploy complex real-time applications. It also separates IMS from the switching core, making it genuinely access independent and specifically allowing interworking with WLAN and similar access technologies.

IMS has also been adopted for use in 3G CDMA networks by the 3rd Generation Partnership Project 2 (3GPP2), but renamed as the Multimedia Domain (MMD). So services based on IMS will interoperate across UMTS and CDMA 3G networks.

ETSI and the International Telecommunication Union (ITU) are now looking to create, among other things, versions of the services provided by a 3GPP IMS core, but extended and adapted to allow delivery to fixed-line terminals, and to enable the partial or complete replacement of the existing core infrastructure supporting these terminals.

Most current (early 2005) IMS services are trials, but wider commercial deployment is likely in 2005. Commercial launches are mostly focused on a single application, especially push-to-talk and video sharing. Most implementations are through a single vendor, although they often involve multiple partnerships, for example for SIP servers, media gateways, media servers, and so on.

The basic idea is that IMS carries signaling and bearer traffic over the IP layer and operates as a "routing engine" or "session control" application that matches user profiles with appropriate call/session-handling servers, and then routes the call or session to the appropriate destination. The architecture includes the capability to add, modify, or delete sessions during an existing multimedia session or circuit-switched call. This opens possibilities of "blended" services that involve simultaneous voice, data, and multimedia sessions.

3GPP TS 23.002 V6.6.0 (2004-12), the most recent technical specification for the 3GPP network architecture, defines IMS as comprising all the core-network elements providing IP multimedia services (such as audio, video, text, chat, etc., and combinations of them) over the packet-switched domain of the core network. The overall network architecture behind this definition has two parts: an access network and a core network. In mobile terms, the access network provides the wireless access points and links to the user, and the core network provides service control and the fixed connectivity to other access points, to other fixed networks, and to service resources, such as databases, interactive announcements, and content delivery.

The core network is assumed also to have two parts, known as domains: a circuit-switched domain and a packet-switched domain. Circuit-switched connections require dedicated network resources to be allocated during a connection – the PSTN is the classic example of a circuit-switched network. Packet-switched connections do not require such dedicated resources, as information is broken down into separate short messages (packets), which are routed independently through the network to their destinations, where they are reassembled into the original information streams. The Internet is the classic example of a packet-switched network.



IMS contains eight basic core-network elements in the packet-switched domain, shown in Figure 1.

These elements are logical functions, not specific devices, and vendors are free to implement the functions as they see fit. The elements are therefore very abstract and involve a lot of obscure acronyms. They are (with very concise descriptions):

There are other essential elements to a complete architecture that either span the core-network circuit-switched and packet-switched domains, or are essential to the mobile system. These include:

Looked at functionally, IMS uses a layered architecture (shown in Figure 2) and comprises a set of interfaces, SIP proxies and servers (such as media servers), and media gateways (for connections to non-IP networks such as the circuit core or the PSTN). A key feature of the layered approach is that call and session control in IMS are independent of the service layer and access network.

The strength of this architecture is that it extends the IP network from user equipment, through the control layers, to the service or called party, while remaining independent of the type of access network. So it works both with legacy networks and new access networks.

It’s worth pointing out that there are actually two SIPs – the Internet Engineering Task Force (IETF) SIP and the NGN/3GPP SIP. And these currently lack interoperability. Although this is an obstacle to practical NGN implementations, there are technologies that will address it, such as Session Border Controllers in the VOIP environment.

Some aspects of IMS are still not really properly defined; for example, the MRFP. This was specified some years ago by the 3GPP to use H.248 (or Megaco) because that was what was available and implementable at the time. So the IMS MRFP has a pretty limited set of functionality – announcements, IVR, simple conferencing, and the like – and is missing the fancier capabilities, like rich conferencing, text to speech, and video, that carriers need for enhanced services.

However, since then, SIP has become the dominant media server control protocol in wireline NGN networks, partly because it is better for use by application servers for providing enhanced services. So it seems possible that wireline networks’ preference for SIP media servers will carry over into wireless networks, resulting in a divergence between the IMS specification and the market’s preference.

There are also issues over the SCIM function, as this acronym hides a potentially complex task, and the IMS specifications are still very skeletal here. The SCIM is going to have to play a key role in some quite complex service packages – and in their associated service interactions – that carriers will devise to meet the needs of different market segments. Although the CSCF provides the primary call control function, the wider tasks of service orchestration and managing integration among legacy applications and services for a better view of the IT and application environment will fall on the SCIM. So much more work needs to be done on the SCIM.

QOS is also a bit of an unknown when it comes to offering complex services that may involve fixed, mobile, legacy, and NGN carriers all on one call. NexTone, for example, argues that the underlying NGN transport layer (basically MPLS/IP) is going to have to become more "session aware," so that it can dynamically determine what applications or services the user is requesting and whether the network transport and network resources are sufficient to service a call. Currently, the company says, the MPLS QOS mechanisms are really aimed at bulk traffic and do not give the fine granularity needed to handle individual calls.

Billing and charging are going to need a lot of coordination and development in a converged multiservices network and environment. Charging orchestration is talked about within the 3GPP specifications to coordinate and manage the CDRs that are created in multiple different places in the network, but there are all sorts of potential complications. For example, in a converged IP/IMS network it is not strictly necessary to charge separately for content and transport, and therefore generate two separate billing records, since the content of IP packets can be known and charged accordingly in a single environment. So there is a real need, to do the job properly, to have some sophisticated level of charging orchestration to manage and coordinate the CDRs and pass them through to the billing system to provide a flexible method of billing and charging.

From the IP/IMS viewpoint, many (if not most) existing networks use legacy technologies – such as TDM and ATM circuit-switching – with user devices that are certainly not IMS-enabled. Although IMS works with circuit-switched networks, its whole point of sophisticated multimedia services is directed towards packet networks and user devices.

Such a massive legacy base creates a headache for a telco’s IMS business case in today’s strict financial environment, because it makes an NGN overlay rollout in incremental stages not very appealing. Telcos need to boost their overall ARPUs significantly and quickly, not just those of the relatively small number of customers that would be using IMS-enabled services under an NGN incremental overlay. Because these networks (both fixed and mobile) are big, with large numbers of customers, this means finding some way of offering at least some of these services to the many non-IMS customers from the beginning.

Fortunately, the high capabilities of IMS offer a potential solution to this problem by supporting an interim stage with an applications layer that can translate at least some of the advanced services offered to the NGN packet users into forms that can be delivered to users of legacy networks, such as the PSTN of PLMN.

Thus media servers could interwork with the PSTN/PLMN by providing a telephone user interface in place of 3G-handset GUIs or Web GUIs. For example, this interface would use text-to-speech translation (and conversely) and IVR to replace some point-and-click operations.Another possibility to ease the business case is to turn the legacy network to advantage by giving the NGN access to existing, well established, enhanced services (such as local number portability and 800 service), rather than recreating these services all over again for the new network. This can be done by using SIP/SS7 gateways in the signaling layer to transfer information between the two domains without having to transfer the calls themselves.

But telcos also need to think about IMS in the context of a full-scale, carrier-class VOIP service – and all this implies – from day one. For example, the IMS-like subsystem must be able to support emergency services. Such requirements may be a tricky proposition for a technology that is not yet proven to provide equivalent levels of service to the existing circuit network, so the transition of mass-market voice service to IMS is likely to be a long-term proposition.

The whole point of IMS is applications and services, and IMS is intended to provide an environment in which it is easier, faster, and cheaper to develop and offer applications and services. It does this through a set of standard devices, interfaces, and elements from which to build such applications and services. Further, rich and dynamic combinations are possible, as IMS allows telcos and service providers to add, modify, or delete sessions or services during a multimedia session or call, and to combine circuit-switched and packet-switched services in the same session. And they can differentiate and bill for a wide range of services and service elements and create new sources of revenue.

Some of the obvious possibilities are

The industry at the moment is particularly excited about the potential of presence, push-to-talk, and video sharing. However, IMS proponents are keen to emphasize that it is not just a matter of entirely new applications and services, as IMS can offer existing applications, such as telephony, but with additional features and capabilities to increase differentiation, usage, and revenues.

A key point also is that IMS will allow telcos and service providers to pick and mix applications to create tailored services for specific customers and industries. An example is the management of major construction sites, where the prime contractor may be employing a large and varying number of subcontractors over the different phases of the project.

So the telco could allow the site engineer, say, to rapidly create and tear down closed user groups (CUGs) of different subcontractors – who may be on different networks – to manage a small group of people working on specific tasks. These CUGs could be coupled with push-to-talk and push-to-see (via videophones), so that the site engineer could quickly see what a new problem was. Presence would show who was on site and available to fix the problem quickly if necessary, and location capabilities would track the plant and equipment needed. Tying these capabilities into backend systems integration would allow the site engineer to raise immediately any authorizations needed – all through a mobile phone.

In principle, IMS should appeal to carriers, telcos, and service providers of all types, wireline and wireless, because it addresses so many of the fundamental issues facing the telecommunications industry as IP, broadband, and mobility redefine what telecom is about.

And it is starting to be deployed. Early examples include Telecom Italia Mobile SpA's (Milan: TIM) November 2004 launch of a video-sharing service over its EDGE and 3G networks, using Nokia Corp.'s (NYSE: NOK) IMS core network products and handset. Another high-profile example is BT Group plc's (NYSE: BT; London: BTA) 21st Century Network project, which will use IMS-style SIP infrastructure as an important part of its new, converged IP network. In the U.S., BellSouth Corp. (NYSE: BLS) says it is deploying SIP-based infrastructure for unicast IP telephony services into a target architecture that would harmonize on the use of SIP for all types of sessions, including IP telephony, multimedia, video-on-demand, and multicast video.

A big attraction of IMS for virtually all kinds of carrier is that it gives application flexibility and a way of diversifying their service portfolios, building larger customer bases, and avoiding the most dreaded of all fates – becoming just bit-pipe providers. It’s a key to extracting further value from the network.

Mobile operators perhaps have the easiest route into IMS because they can use it in a fairly simple way at first to support a handful of key mobile capabilities, such as IM and presence management, that they know they must have. And since other desirable services, such as POC, are becoming available in IMS form, it seems sensible to start installing IMS-compliant systems, rather than proprietary ones.

Large telcos are probably the players with the most at stake over IMS simply because their networks and user bases are so big, and IMS improves their ability to manage these assets and to extract and retain value from them. In particular, telcos with both wireline and wireless networks are a big target for IMS because it allows them to converge the two into a cohesive single network over which they can deliver, in principle, any real-time application.

It also allows them to do almost the exact opposite, and to split their networks open in very interesting ways. So a carrier could support multiple MVNOs, for example, by opening its control layer to their third-party applications layers. Or the carrier could host the MVNO application servers within its own service-layer facilities.

But IMS will also significantly sharpen the competitive threat to large telcos because it will enable service providers of all types, even those without networks, to offer services that may be a closer fit to the needs of the telcos’ own customers. Telcos are really going to have to understand the business models and the areas of the communications they work in if they are to retain business in a converged service environment.

3GPP: 3G Partnership Project
AS: Application Server
BGCF: Breakout Gateway Control Function
CDMA: Code Division Multiple Access
CS: Circuit Switched
CSCF: Call Session Control Function
CUG: Closed User Group
DSL: Digital Subscriber Line
GPRS: General Packet Radio Service
GSM: Global System for Mobile communications
GUI: Graphical User Interface
HSS: Home Subscriber Server
IETF: Internet Engineering Task Force
IMS: IP Multimedia Subsystem
IMS-MGW: IP Multimedia Subsystem – Media Gateway Function
IP: Internet Protocol
ISP: Internet Service Provider
IVR: Interactive Voice Response
LAN: Local Area Network
MAC: Medium Access Protocol
MGCF: Media Gateway Control Function
MRFC: Multimedia Resource Function Controller
MRFP: Multimedia Resource Function Processor
MS: Mobile Station
MVNO: Mobile Virtual Network Operator
PDF: Policy Decision Function
POC: Push-to-talk Over Cellular
PS: Packet Switched
PLMN: Public Land Mobile Network
PSTN: Public Switched Telephone Network
QOS: Quality of Service
RAN: Radio Access Network
SGF: Signaling Gateway Function
SIM: Subscriber Identity Module
SIP: Session Initiation Protocol
SLP: Subscription Locator Function
UE: User Equipment
UMTS: Universal Mobile Telecommunications System
VOIP: Voice Over IP
WLAN: Wireless Local Area Network