Mobility Pattern Free Dynamic and Effective Location Update in Manet

The beaconing approach is the key function in geographic routing to disseminate the location. However, the node mobility is a prominent challenge to the beacon based location broadcasting schemes resulting in high routing overhead. The conventional methods allow some errors on location prediction. As a result, the mobile nodes update their location when the predicted location exceeds the allowable error range. However, the prediction error is more sensible for boundary nodes than adjacent nodes, as the boundary nodes located in the proximity area act as greedy nodes. Consequently, allowing the static prediction-error for all nodes does not efficiently reduce the overhead while maintaining the neighbor list accuracy. To deal with these issues, this work proposes a system called "Mobility pattern free Dynamic and Effective Location update" (MODEL) for the maintenance of the trade-off between overhead and precision. Instead of allowing the static prediction-error, the Dynamic Acceptance Error Rate (DAR) in MODEL dynamically calculates the error range to the boundary and adjacent nodes and enhances the neighbor list accuracy with routing overhead. Due to the sensitivity of boundary nodes to the location being accurate, the MODEL efficiently exploits the fuzzy algorithm to allow a minimum error in predicting location rather than in adjacent nodes. This work simulates the proposed MODEL in NS2 simulator and compares the performance of the existing Load Balanced-Dynamic Beaconing Greedy Perimeter Stateless Routing (LB-DB-GPSR).