Novel MAP selection scheme using location history in Hierarchical MIPv6 networks

In this paper, a novel Mobility Anchor Point (MAP) selection scheme based on location history is proposed for Hierarchical MIPv6 networks. Considering Mobile Nodes (MNs) moving fast but only within a limited region, a new parameter, called location rate, is defined in our scheme to represent the number of different access routers the MN has previously attached to in unit time on average. A compound variable, consisting of the location rate and some other MN-related parameters, is derived and used to compare with a group of network-related thresholds, so that MAPs with minimum total costs (containing binding update cost and packet tunneling cost) for different MNs could be selected. Simulations with a modified random walk mobility model show that our scheme has good and steady performance for selecting low-cost MAPs in different situations, while other schemes (e.g. furthest, nearest, mobilitybased and session-to-mobility ratio based) are only suitable for certain cases. Besides, the use of the compound variable and the group of thresholds simplifies the selection procedure, so MNs don't need to do any complex calculation of MAPs' total costs.