Optimizing tunnel management in predictive handover protocols

Predictive handover protocols enable seamless handover of mobile nodes with reduced packet loss. The seamlessness is achieved due to the establishment and maintenance of a bi-directional tunnel between the mobile node’s previous IP subnet and new IP subnet over backhaul links. During the time the mobile node is executing a handover, all the packets destined for the mobile node will get forwarded over this tunnel to the new IP subnet where they will get buffered till the MN becomes IP capable on the new subnet.Although this strategy will effectively reduce the packet losses during handover execution but, in view of the increasing number of mobile users and the ever increasing traffic volume, it will incur a high tunneling load over bandwidth limited backhaul link between the previous and the new subnet leading to several performance issues. It has also been observed that the performance efficiency of predictive handover protocols is dependent on the speed of the mobile node and also on the timing of initiating the handover process. This dependence also has a direct bearing on the tunneling load, and hence the link utilization.This paper explores the overall performance implications of a high tunneling load with reference to the predictive handover mode of the standard FMIPv6 protocol, and will discuss in detail a proposed solution called Proactive Bindings for FMIPv6 (PB-FMIPv6), which extends and optimizes in particular the FMIPv6 protocol; and in general any predictive handover protocol that utilizes a forwarding tunnel between previous and next IP subnets.