A MAP-Controlled Load Balance Scheme for Hierarchical Mobile IPv6

Abstract

In Hierarchical Mobile IPv6 (HMIPv6) networks, the Mobility Anchor Point (MAP) handles both intra-domain handovers and tunneling traffic for Mobile Nodes (MNs), which is prone to be a single point of bottleneck. In order to increase reliability and flexibility, multiple MAPs are deployed in a management domain. In this case, load balance among MAPs should be carefully considered. Based on the fact that the MAP can react to the change of its traffic load more quickly than the MN, a MAP-Controlled Load Balance scheme (MCLB) is proposed. MCLB consists of two algorithms, i.e. probability-based MAP selection algorithm and MAP-controlled adjusting algorithm. When a MN enters a management domain, it can register with more than one MAP by using the probability-based MAP selection algorithm, which is based on the probability calculated from the priority of MAPs besides the properties of the MN. To make MAPs in the network remain in the steady state, a Markov chain based model is used to induce the probability in the algorithm. Moreover, when the traffic load at a MAP trends to be higher, MAP performs the MAP-controlled adjusting algorithm to select and redirect some MNs to other MAPs with lower traffic load. Experimental results from a real test-bed show that MCLB can achieve load balance among MAPs with low overhead.