Wormhole attack detection and recovery for secure range free localization in large-scale wireless sensor networks

Usually, resource-constrained sensors operate unattended in an infrastructure-less environment. It attracts various active and passive attacks where wormhole attack gains prime attention due to its effortless implementation in comparison to its devastating effect on localization schemes. Further, an implicit wormhole with a single tunnel makes it difficult to uncover it. A wormhole disturbs hop counts erratically among the sensor pairs which in turn destroys most of the localization algorithms like DV-Hop and its various successors. Therefore, an algorithm is required to detect anomalous sensors as outliers and remove them from participating as reference points in localization. Thus, in this paper, a secure optimized localization in large-scale WSN (SOLLW) is proposed where outliers are detected and removed by using a one-class support vector machine. Further, SOLLW recovers location through linear optimization by introducing error factors in distance values between every sensor pair. The simulation validates SOLLW in comparison to other algorithms.