Hop-Count based Distance Estimation in Wireless Sensor Networks under Byzantine Attacks

Abstract

Estimation of the Euclidean distance between a source and a sink sensor by using the hop-count of that sensor and its neighbor is an interesting research problem in Wireless Sensor Network. In some cases, the distance is wrongly estimated due to error in the value of hop-count. One such case is the case of Byzantine attack where an outsider tries to attack some sensors and change its hop-count to degrade the performance of the network in distance estimation. In this work, we have analyzed the effect of Byzantine attack in the method proposed previously. We have considered three previous proposed methods: Minimum Mean Square Error (MMSE) estimator, Expected per hop progress (EPHP) estimator and Neighbor partition (DNP) estimator. Then we propose a modified source to sensor distance estimation method in which we first correct the hop-count data based on the observations of sensor’s hop-count and the information of its neighbors’ hop-count. Simulation results show that the proposed method outperforms other well-known hop-count based source-sensor distance estimation methods.