AbstractAs opposed to mobile next generation network (NGN) approach aiming at just carrying circuit services on top of the Internet protocol (IP), the IP multimedia subsystem (IMS) offers operators the opportunity to build an open IP based service infrastructure that will enable an easy deployment of new rich multimedia communication services mixing telecom and data services. Before trying to emulate circuit switched domain, IMS will first provide new services (such as instant messaging, presence, push-to-talk over cellular [PoC]) that are not too demanding for the underlying radio access network. These services will attract customers and create new telecommunication needs without inducing the technical risks and costs of speech with stringent quality over packet radio.
IntroductionIMS is a new framework, basically specified for mobile networks, for providing IP telecommunication services. It has been introduced by the third generation partnership project (3GPP) in two phases (Release 5 and Release 6) for universal mobile telecommunications system (UMTS) networks. An IP multimedia framework was later introduced by 3GPP2 as the multimedia domain (MMD) for third generation code division multiple access 2000 (CDMA2000) networks, and finally harmonized with IMS. Real-time services can only be properly supported using the Release 6 IMS specifications. Currently, a Release 7 is being worked out by 3GPP taking into account the requirements from Telecoms and Internet-Converged Services and Protocols for Advanced Networking (TISPAN) and European Telecommunications Standard Institute (ETSI) on fixed IMS and NGN requirements.
The IMS concept was introduced to address the following network and user requirements:
* Deliver person-to-person real-time IP-based multimedia communications (e.g. voice or videotelephony) as well as person-to-machine communications (e.g. gaming service).
* Fully integrate real-time with non-real-time multimedia communications (e.g. live streaming and chat).
* Enable different services and applications to interact (e.g. combined use of presence and instant messaging).
* Easy user setup of multiple services in a single session or multiple simultaneous synchronized sessions.
3GPP(2) defines a reference architecture, functional splits, protocol specifications, use cases, and so on, for the IMS phases in the context of packet-based UMTS (CDMA2000) access. Many of the features, such as quality of service (QoS) control, full security and full roaming, need specific features to be available in the radio access and/or core network. Other features, such as the session initiation protocol (SIP) session control, integrated access and charging for combined applications, open up the integration of third-party applications. They can be provided today on pre-IMS packet access networks, such as the general packet radio service (GPRS), enhanced data rates for Global System of Mobile Communications (GSM) evolution (EDGE), and CDMA 1x as well as fixed NGN based packet access.
IMS Market Trends and Business ChallengeGlobal carriers are facing decreasing average revenues per user (ARPU). Due to strong competition, the prices of voice calls and Internet access are declining and the carriers' margins are being reduced. Carriers can no longer afford to offer only voice and ever faster Internet access because these items are becoming mere commodities. They need to identify attractive services that can be bundled with their basic access offer in order to reduce churn.
Most carriers have started looking for such services. Mobile networks operators have introduced portals (CDMA, i-modeTM, and GPRS) with the aim of offering content on mobile phones. Initial attempts were met with varying degrees of success: i-mode was successful in Japan, while the wireless application protocol (WAP) proved to be unsuccessful in Western Europe.
Besides e-mail, most of these portal services are person-to-server services in which a user interacts with a content server. One of the drawbacks of such services is that carriers have to obtain content, which means entering a new business field and negotiating with content providers. This process is costly and undermines the anticipated revenue opportunities from portal services.
Today, however, mobile network operators derive the bulk of their revenues from person-to-person services, mainly voice and short message service (SMS). The core of their business is to enable two users to communicate. Content is created by the users, so wouldn't it make sense to look for new person-to-person services?
E-mail and instant messaging are proving to be successful mobile services, as they already are on the wireline Internet. The multimedia messaging service (MMS), which enables mobile users to exchange pictures, is an example of carriers launching a new person-to-person service.
One area that mobile carriers have not yet been able to address is in real-time rich multimedia person-to-person communication. Video telephony over a UMTS network is a first step in this direction; the communication is enriched with a real-time video component. A UMTS videophone enables you to see the person to whom you are talking, and vice versa. However, because it is based on circuit switching, UMTS video telephony suffers some limitations; for example, a chat component cannot be added to a video telephone call.
Mobile carriers cannot offer their customers the freedom to mix multimedia components (text, pictures, audio, voice, video) within one call. Nowadays, a two party voice call cannot be extended to a multi-party audio and video conference. IMS overcomes such limitations and makes the previous scenario possible.
Moreover, in the future, it will make sense for operators to bridge the gap between traditional telecom services (voice call, SMS) and data services (e-mail, browsing, and instant messaging). Why would not it be possible to start a video call from an Internet directory, extend a text chat to a voice call, show a web and WAP page while making a video call, or show a photograph (provided by the caller) instead of a numerical caller identifier? Mobile carriers will make their offer more attractive if they can provide their customers with the right blend of telecommunication and data services seamlessly interwoven. Because of its IP foundations and the extensibility of SIP, along with the IMS signaling protocol, IMS enables telecommunication and data services to be mixed.
Real-time rich multimedia communication mixing telecom and data services has started to happen in wireline broadband networks. The major instant messaging providers are offering a mix of e-mail, text chat, multimedia file exchange, voice and video calls, and video chat. They have been successful in attracting millions of young people to this service mix, who are in turn tomorrow's adult consumers. Mobile carriers will have to offer them the same multimedia rich environment if they want to prevail over the pure information technology (IT) players, like Microsoft MSN, Yahoo, and America Online (AOL).
However, mobile carriers will have to choose whether they want to become IP pipe providers or IP telco providers. On the one hand, if they become "IP pipe providers", they will focus on transporting IP packets while the value-added services will be offered by others. This is similar to the situation in fixed networks in which broadband access is offered by the carriers, while the most successful IP services (e-mail, instant messaging, peer-to-peer file exchange) are offered by application service providers (ASP). The risk is that network access might become a commodity, thus leaving only small margins for the carrier. On the other hand, if the mobile carriers can become "IP telco providers" they will retain a greater part from the value chain, not only network access but also applications and services, payment, service brokering and even content. The portals and associated payment facilities (money charged for the application service providers and sharing the revenue with them) are a first step in this direction. IMS is a further step because it is carrier-centric, thus enabling them to retain control over rich multimedia sessions.
Evolution toward Work-Style ServicesThe integration of different media opens up new possibilities for far richer services than those available today and for which users will probably be willing to pay.
IMS is concerned with the creation and deployment of multimedia telecommunication services over any IP network. Most importantly, this includes person-to-person real-time services (such as voice) over packet switched (PS) networks, removing the need for a circuit switched (CS) domain. However, emulation of mobile CS services is not the primary goal of IMS, although their replacement by IMS services is unavoidable in the long run. Furthermore, it will be necessary for IMS to interwork with external CS networks, such as the public switched telephone network (PSTN) or public land mobile network (PLMN), even if there is no longer a circuit switched core network (CN) domain.
The main argument for introducing IMS is to support creative services, for example, from IT-oriented third parties.
IMS brings the power and wealth of Internet services to the mobile environment to a far greater extent than is possible today. In effect, IMS takes the best from the switched and packet worlds. In this context, a more controlled value chain is possible which is of greater benefit to the operators than are current Internet business models. For mobile operators, such models should continue to persist since users will certainly be willing to pay for the combination of rich Internet services and mobility.
IMS enables sessions to be established between multiple users and devices, and allows multiple services to be carried on a single bearer channel. Examples are the ability to set up:
* Separate services within a single session; these may be synchronized (e.g. voice and video for video telephony) or not (e.g. video and chat).
* Multiple simultaneous single sessions of unrelated services (e.g. voice call in parallel with a presence session).
* Easy switching between point-to-point and point-to-multipoint person to person sessions, without the burden of pre-programmed conference services on the PSTN. Combined with the flexibility of packet networks, this will allow walkie-talkie like services, such as PoC, in which a user simply has to push a button to send voice messages to a group of peers.
* Ease of simultaneously ringing a called users different terminals.
The service combination also guarantees session completion, which does not depend on the called party having a compatible terminal. For instance, a session can be accepted without the video component or the video component can be routed to another destination. Together all these features enable IMS to provide a significantly enhanced user experience and promises to be the basis for radically new services.
IMS is designed to be agnostic to the underlying access network, although initially it has been designed to run over mobile packet networks (e.g. GPRS, EDGE, UMTS, CDMA-based). In a second phase, it will also run on fixed (wireless) broadband networks (e.g. x digital subscriber line (xDSL), wireless local area network [WLAN], and cable). Consequently, IMS holds out the promise of seamlessly supporting converged services on fixed and mobile networks.
Safe Deployment of IMSAlthough IMS is access agnostic, initially it will be deployed in mobile networks. Consequently, taking an end-to-end service perspective, mobility and particularly radio constraints must be taken into account. However, some other network-related constraints are also relevant, although not necessarily linked to mobile networks.
The use of IPv6 for IMS — as required by 3GPP — makes it necessary for an operator to deploy almost entirely new network layers at the same time as rolling out IMS:
* New service layer elements to provide new IMS services
* Totally new session layer, that is, a new networking layer based on SIP
* New transport layer based on IPv6
* Associated back-office features, such as security, routing and charging. Since some IMS services are totally new they might require new charging concepts and new charging information to be stored in the relevant network nodes.
Providing initial support for IPv4 over IMS overcomes the challenge of deploying IMS services over a new transport layer. Furthermore, deploying IMS solely on IPv6 may be an issue for communication services provided by the mobile operator to enterprise networks, which might be deploying IPv6 at their own pace.
IMS is indeed a good tool to deliver the same set of services to both wire-line and wireless users. For this reason, it might prove beneficial to deploy IMS over IPv4, at least during the roll-out phase. Nevertheless, the operator should not wait too long to carry out the necessary migration to IPv6 in order to avoid high migration costs when IMS has become established. The efficient transport of voice over a radio interface is another challenge that must be met when delivering real-time conversational services (such as VoIP over IMS). As the radio is the most expensive part of a PLMN, it must be optimized, otherwise the reduction capital expenditures (CAPEX) and operational expenditures (OPEX) potentially brought by IMS in the CN would be ruined.
A first step in providing efficient transmission of conversational voice over radio has been provided by robust header compression (RoHC), but further improvements are still needed. Answers to various questions, including the following, have to be provided before deploying IMS services that supersede the CS domain services:
* How is it possible to cope with variations in the length of voice frames over radio as a result of RoHC compression?
* How can one deal with real-time control protocol (RTCP) considering that its traffic is very sporadic but may correspond to frames with very different lengths which might delay real-time protocol (RTP) traffic?
* How to optimize the transfer of adaptive multi-rate (AMR) speech over the packet domain using techniques such as AMR codec mode control and unequal error protection (UEP)?
It is important to point out that the level of end-to-end QoS supported by the infrastructure determines the nature of the services that can be supported (e.g. real-time or non-real-time). In the case of real-time mobile services over IMS, the GPRS and UMTS infrastructure has to meet hard QoS constraints, for instance, full support for conversational quality. Such constraints cannot be met by today's network, for example, because of the lack of handover support for 2.5G packet radio.
ConclusionIMS can provide new person-to-person multimedia communication services going beyond those available on today's 3G networks. Because IMS is IP based, it will blend telecommunication and data services, keeping the best of both the circuit-switched and packet worlds. As IMS was designed for network operators, it keeps multimedia call control with the mobile carriers, who can thus win a bigger part of the value chain (access, application and services, brokering, payment, etc.) and can become IP telco providers instead of "IP pipe providers".
Labels: Technology
