Distributed Target Localization Estimation Methods Under Different Byzantine Attacks

Distributed target localizaiton reserches have been hotspots based on the development of wireless sensor networks(WSNs). Due to the open network environment, the distributed target localization is vulnerable to be attacked by malicious nodes. It is extremely important to obtain the target location safely and effectively. Because of the high computational complexity, low fault-tolerance and the simple probability of Byzantine attacks for the existing distributed target localization methods. In this paper, we mainly carry out research depend on distributed target localization estimation methods under different Byzantine attacks.To the limitations of the existing distributed target localization methods, a block iteration distribution localization method with the region extension factor is proposed. The proposed method overcomes the influence when the target is located in the border of the sub-regions, and improves the ability of fault tolerance. The estimate target location is the centre of the reserved region which reduces the computational complexity effectively. In addition, the error probability, probability of region decision (PRD), root mean square error(RMSE) and communication costs are proposed to analyze the localization performance. We also analyze the localization performance for single probability Byzantine attack. Theoretical analysis and simulation results show that the proposed method achieves a better performance compared with the optimal Byzantine attack.To the problem for simple probability of Byzantine attack, a complex Byzantine attack model is proposed. The strategies of Byzantine attack are analyzed for the independent attack and the collaborative attack. For the independent Byzantine attack, the fraction of Byzantine nodes to blind the Fusion Center(FC) is obtained. We also analyze the localization performance under the resource-constrained Byzantine attack. For the collaborative Byzantine attack, a heuristic method is proposed to obtain the fraction of Byzantine nodes to blind the FC. We also analyze the optimal collaborative Byzantine attack strategy and the influence on the localization performance. Due to the intercommunication, the localization performance of the collaborative Byzantine attack is decrease compared with the independent attack.