No abstract

No abstract

No abstract

SUMMARYTarget tracking in wireless sensor networks is a well-known application. In real life scenario, target mobility can be predicted using well-known filters. In this paper, we explain an approach to model the pattern of movement of a target on the basis of target data available. This method utilizes filter techniques to predict the target and a curve-fitting algorithm to model the mobility of a target in both linear and non-linear motion patterns. Two alternate strategies to achieve mobility approximation have been proposed and compared. The efficacy of the algorithm is, further, adjudged by comparing its mobility prediction vis-a-vis the Kalman filter. Simulation results show that with sufficient data, the mobility pattern of the target can be fairly calculated even if the target moves unpredictably.

No abstract