An Attack-Resistant Localization Algorithm with Adaptive Weights for Range-Based Wireless Sensor Networks

Abstract Localization is a crucial problem for robot control in the wireless sensors network with malicious nodes. In this paper, an attack-resistant localization algorithm with adaptive weight is proposed to achieve an accurate localization without any prior information of malicious anchors. The core idea of the proposed algorithm is adjusting the weights of malicious anchors based on the estimation for the attack value of malicious anchors to eliminate the influence of malicious anchors. Meanwhile, the subfunction's gradient is used to avoid the local minimum of non-convex function based on the particularity of the positioning problem. Three kinds of simulations are conducted to illustrate that the proposed algorithm can achieve an accurate localization in different types of attacks. Furthermore, two comparisons with other two localization methods are presented to show the advantage of the proposed algorithm by simulations and experiments.