Predictive Mobility Management with Delay Optimizations in 802.11 Infrastructure Networks

In 802.11 wireless infrastructure networks, as the mobile node moves from the current access point to another, the active connections will not be badly dropped if the handoff is smooth and if there are sufficient resources reserved in the target access point. The predictive mobility management scheme we propose has primarily - mobility prediction block, delay management block and resource management block that aids the handoff. In a symmetric grid of access points, within a grid of regions, by location tracking and data mining, we predict the mobility pattern of mobile node with good accuracy. Active pre-scanning of mobile nodes, pre-authenticating neighbouring access points and pre-reassociation using mobility prediction are used to reduce the probe delay and authentication delay and reassociation delay respectively. The model implements reservation in two stages by using mobility prediction results and traffic type, so that sufficient resources can be reserved when the mobile node does the handoff. The overall mobility management scheme thus improves the quality of service and enables smooth handoff. Elaborate performance simulation is done in Java to verify the proposed model