Improving machine to machine communication and ims load balancing in 4g networks

The fourth Generation (4G) wireless networks are designed to accommodate the fast growth in mobile data traffic. The high data rate and all IP design opened the door for the introduction of new services. These services are putting more pressure on the network performance and resources. Voice calls, the basic service in wireless networks, are supported by a new overlay network, the IP Multimedia Subsystem (IMS). Another service that is expected to undergo fast growth is Machine to Machine (M2M) communication over wireless networks. This dissertation addresses the challenges of supporting a large number of M2M terminals in the 4G wireless networks and the challenges of load balancing in the IMS control plane. The 4G wireless networks are designed for User Equipment (UE) used in Human to Human (H2H) communication. This poses some deployment issues for M2M communication over the 4G wireless networks. The existing numbering scheme, for example, is not suitable for identifying the large number of M2M devices. The dissertation discusses the addressing of M2M terminals in the 4G wireless networks and proposes a solution to double the address space to support up to 1015 new globally unique identities for M2M terminals. Management of the subscription of M2M devices is one of the major challenges to M2M operators. The existing semi-manual process can be very labor-intensive, and thus costly, for large number of M2M devices such as utility meters. This dissertation addresses the challenges of M2M terminals subscription management and proposes an over the air scheme to remotely manage M2M subscription. The proposal supports initial subscription activation and subscription change to a new network operator. The large number of M2M devices can easily degrade the human users Quality of Experience (QoE) without careful scheduling of the radio resources. This dissertation proposes an M2M-aware LTE scheduler that preserves the human users' QoE and provides fair scheduling of M2M terminals. The two main entities in the IMS signaling plane are the Proxy Call Session Control Function (P-CSCF) and Serving Call Session Control Function (S-CSCF). Both entities are assigned to end users at registration for the duration of their connection to the IMS network. The assignment of end users to the two nodes does not take the node's current load into consideration as a selection criterion. This can lead to driving some of the nodes into overload while other similar nodes are still under-utilized. A cooperative framework is proposed for load balancing in the IMS network signaling plane to evenly distribute the users between available nodes.