Projection-Based Consensus Algorithm for Secure Localization in Ultra-Wideband Sensor Networks

Abstract

Secure localization in sensor networks is a promising research in engineering field. This paper considers a scene that there exists malicious anchor in the sensor network composed by ultra-wideband (UWB) devices. A projection based consensus localization algorithm is proposed to overcome the effect of malicious anchors and achieve precise positioning for tag nodes. The algorithm firstly adopts the constrained consensus method to control the virtual agents until they stop move. Then, a selection strategy based on median is used to choose appropriate virtual agents and calculate the mean of their positions. At last, the gradient descent method is used to achieve localization with only considering the measurements from anchors corresponding to the chosen virtual agents in the previous step and setting the mean position of the chosen virtual agents as the initial solution. Furthermore, three simulations are conducted to illustrate the validity of the proposed algorithm. Compared with the traditional method, the proposed algorithm can locate the tag successfully when there exists malicious anchor in the environment.