Change-Point Monitoring for Secure In-Network Aggregation in Wireless Sensor Networks

Abstract

Secure in-network aggregation in wireless sensor networks is a necessary and challenging task. In this paper, based on an extended Kalman filter which can facilitate us to set up a normal range of the neighbor's future transmitted aggregated values, we further apply an algorithm of combining cumulative summation and generalized likelihood ratio, which can utilize the cumulative sum of the deviations between measured values and estimated values, and derive a normal range which is more sensitive to attacks. We conduct experiments and simulations to evaluate our proposed local detection mechanisms under different aggregation functions.