Secure Localization Based on Gradient Descent Approach Under Noncoordinated and Coordinated Attacks in WSNs

In many applications of wireless sensor networks, localization is one of the important issues that affect the quality of network operation. Most localization algorithms rely on anchor nodes that their positions are known and they help positioning of the remaining nodes. When the anchor nodes are affected by the attacker, they send false information in the network. Thus, it is required to design a secure localization algorithm that removes the effect of the false information. In this paper, we describe a secure localization algorithm based on the gradient descent approach with pruning step and explain the noncoordinated, coordinated attacks and the combination of them as well. We evaluate the performance of the algorithm under the combination of attacks. Then, to reduce the localization error, we propose a new method by considering the direction of force vectors in pruning step. Simulation results show the effectiveness of the proposed method in the hostile environment.