Self organizing IP Multimedia Subsystem

Abstract

While there have been tremendous efforts to develop the architecture and protocols to support advanced Internet-based services over 3G and 4G networks, IMS is far from being deployed in wide scale. Effort to create an operator controlled signaling infrastructure using IP-based protocols has resulted in a large number of functional components and interactions among those components. Thus, the carriers are trying to explore alternative ways to deploy IMS that will allow them to manage their network in a cost effective manner while offering the value-added services. One of such approaches is self-organization of IMS. The self-organizing IMS can enable the IMS functional components and corresponding nodes to adapt them dynamically based on the features like network load, number of users and available system resources. This chapter introduces such a self-organizing and adaptive IMS architecture, describes the advanced functions and demonstrates the initial results from the prototype test-bed. In particular, we show how all IMS functional components can be merged and split among different nodes as the network demand and environment change without disrupting the ongoing sessions or calls. Although it is too early to conclude the effectiveness of self-organizing IMS, initial results

User-transparent reconfiguration method for self-organizing IP multimedia subsystem

“…On the basis of supplied information, the master node assigns IMS components to nodes and gives instructions to them, and the nodes make the required changes. For these behaviors, additional features such as node discovery and role request and assignment, and interaction protocols between the nodes and the master node have been proposed in [3]. In addition, after the IMS configuration is changed, those IMS components affected at each node need to take care of active sessions that were established before the reconfiguration.…”

“…The former tries to inform the UE of the IMS reconfiguration and make them responsive to the change by using standardized protocols for session handoff. For this approach, we proposed an efficient method to notify the affected UE of the change without causing a major impact on the existing IMS [3,4]. However, this method requires some modifications to the UE as well as to the P-CSCF and S-CSCF.…”

“…If we use a backup server, we need to consider an efficient backup method for sessions. In order to back up and transfer the session state, we also propose an efficient way using a mutual complementarity mechanism between CSCFs because the P-CSCF and S-CSCF have almost same session information [3].…”

“…Thus, this self-organizing mechanism also needs to support service continuity to allow the UEs to continue their services seamlessly after the reconfiguration. In our prior work [3,4] the self-organizing concept has been defined for efficient IMS deployment. However, the scheme proposed in this aforementioned paper requires the involvement of UEs when the IMS core network reconfiguration happens, leading to significant signaling overhead and modification of UEs' protocol stack.…”

Service continuity support in self-organizing IMS networks

“…Furthermore, these papers did not take into account reconfiguration of the functions or roles, for example due to nodes failure or for load balancing. The work presented in [1], namely UE involved method, is the first truly self-organization concept applied to IMS networks without using overlay networks. This paper works around the concept of self-organization for IMS networks and the main focus is the UE non involved case.…”