A Critical Evaluation of the IMS Presence Service Ling Lin 1 , Antonio Liotta 1 Abstract The IP Multimedia Subsystem (IMS) defined by the 3rd Generation Partnership Projects (3GPP) is a standardized architecture for the Next Generation Network. In this paper, we provide an insight into the presence service, one of the fundamental building blocks of the IMS. Our prototype-based study is unique of its kind and helps identifying the factors which limit the scalability of the current version of the presence service, which will in turn dramatically limit the performance of advanced IMS services relying on presence information. We argue that the client-server paradigm behind the current IMS architecture does not suite the requirements of the IMS system, which defies the very purpose of its introduction. We finally elaborate on possible avenues for addressing this problem. 1. Introduction The IMS, standardized by 3GPP, specifies a service-centric framework for all-IP networks, covering both wired and wireless networks. Based on SIP, the IMS provides an open interface for third party services to be easily integrated and an opportunity for end-users to experience rich and personalized services. The IMS architecture will deploy both current and future Internet services, such as instant message, video conference, online game etc. The Presence Service is one of basic services in IMS that are likely to become omnipresent, to provide extensive customized information for both end users and third party services. The Presence Service provides the ability to manage presence information of a user's device, service or service media even whilst roaming. [7] In this article, we provide an insight into the performance of the IMS presence service. We have developed an IMS prototype system, including x-CSCF, presence server and instant message application (see Section 2 background information on the IMS). Our study indicates that, despite being such an important building block which has already been standardized, the presence service will be a burden in the IMS system. Our prototype-based study is unique of its kind and helps identifying the factors which limit the scalability of the current version of the presence service, which will in turn dramatically limit the performance of advanced IMS services relying on presence information. We argue that the client-server paradigm behind the current IMS architecture does not suite the requirements of the IMS, which defies the very purpose of its introduction. We finally elaborate on possible avenues for addressing this problem. For the benefit of those readers who are not familiar with the IMS, Session 2 provides a brief description of its architecture, particularly the presence service. Section 3 explains the IMS prototype design and implementation. Section 4 describes the assessment of the IMS presence service, while in Section 5 we elaborate on how its shortcomings may be addressed. Conclusions are drawn in Section 6. 1 Department of Electronic Systems Engineering, University of Essex, Wivenhoe Park, Colchester. United Kingdom. Email: {llini | aliotta}@essex.ac.uk 19 Proceedings of MoMM2006