Position Prediction based Multicast Routing (PPMR) using Kalman Filter over VANET

We proposed a Position Prediction based Multicast Routing (PPMR) for VANET with the objectives of alleviating the broadcast storm problem of multicast tree discovery and simultaneously minimizing the number of forwarding vehicles. A unicast protocol, Position Prediction based Unicast Routing (PPUR) serving as a base for our multicast protocol is also proposed. With the help of Global Positioning System (GPS) device, Inertial Navigation System (INS) and digital map installed on vehicles the proposed protocol gathers the position's information of vehicles forming the network. The position and mobility information are available at each of the destination vehicles for every source-destination route pairs helping destination vehicles to make the topology of the network for path prediction. When a route breaks then the destination node tries to inform the source about other predicted routes. This prediction done by the destination vehicle helps the data source vehicle to transfer data without any route discovery and thus eliminating the broadcast-storm problem with minimizing the route discovery latency also. We are using Kalman Filter (KF) and its modifications to remove the inaccuracies and fusion of data collected from the GPS and INS. The values from KF is used for finding distances between vehicles under the assumption of spherical earth being projected to a flat surface.