Performance optimisations in UMTS packet switched call control

Abstract

Universal Mobile Telecommunications System (UMTS) uses the session initiation protocol (SIP) for packet switched (PS) call control (CC). Recent publications have shown that call set-up delay of PS CC varies between five and seven seconds, depending on the compression of the signalling traffic. The mentioned call set-up delay is a critical factor in service acceptance. Another important issue is the traffic load of CC. We have investigated the procedures of PS CC in UMTS in order to optimise signalling traffic load and call set-up delay. We assume that there are always enough resources in the PS core networks of UMTS. We worked out two different architectures that allow optimising either call set-up delay or signalling traffic load. If we perform radio access network (RAN) resource allocation and media capability negotiation in parallel, it is possible to reduce the call set-up delay by a third. The segmented status type of the SIP preconditions extension applied to UMTS allows us to reduce signalling traffic in UMTS significantly. To illustrate the potential of this approach, it is essential to state that this can be combined with every compression algorithm, which can additionally save bandwidth, e.g. in the RAN. The paper introduces the two new approaches and compares them with conventional UMTS PS CC. We discuss pros and cons of the new interactions and provide calculations of estimated performance gains.