Secure Localization Algorithms Against Localization Attacks in Wireless Sensor Networks

The accuracy of the location estimation of sensors in wireless sensor networks (WSNs) is important to many applications such as environment surveillance applications. In large-scale and distributed WSNs, locations of the sensors may not be recorded during the deployment process, resulting in the need of localization techniques to help collect the sensor location information. WSNs are highly vulnerable to localization attacks since the localization techniques totally rely on the neighboring relations to determine the location of sensors. In other words, some of malicious sensor nodes can take the advantage of the neighboring relations to manipulate the localization process. In this paper, we propose two localization algorithms to overcome both common types of localization attacks, independent attacks and collusion attacks. In the first algorithm, we propose the Improved Randomized Consistency Position Algorithm to determine the location of unknown nodes using the location estimation information provided by the three random anchor nodes and using the PSO algorithm paradigm to identify the coordinates of the unknown nodes' location. In addition, the second algorithm, which is a combination of voting method, location optimization, and the PSO algorithm paradigm to identify the location of unknown nodes, is referred to as the Enhanced Attack-Resistant Secure Localization Algorithm. Through rigorous theoretical analysis and experimentation, we demonstrate that our proposed localization algorithms yield better performance comparing to the existing methods.