Probabilistic Evidence Aggregation for Malicious Node Position Bounding in Wireless Networks

Hyperbolic position bounding of malicious devices aims to estimate the location of a wireless network rogue insider that transmits an attack message containing falsified information to mislead honest nodes. A probabilistic path loss model is used to construct an area in Euclidian space bounded by minimum and maximum distance difference hyperbolas between each pair of trusted receivers. This hyperbolic area is said to contain the rogue insider with a degree of confidence. We explore the combination of evidence provided by a set of multiple receiver pairs supporting the intersection of their hyperbolic space. We propose a novel heuristic scheme to aggregate area probability so that the combined degree of confidence ascribed to the intersecting space is computed according to a paradigm of supportive rather than competitive evidence. Performance evaluation concludes that our aggregation model yields a probability distribution better fitted to experimental location estimation results than a redistributive paradigm.