Source Location Privacy-Aware Data Aggregation Scheduling for Wireless Sensor Networks

Source location privacy (SLP) is an important property for the class of asset monitoring problems in wireless sensor networks (WSNs). SLP aims to prevent an attacker from finding a valuable asset when a WSN node is broadcasting information due to the detection of the asset. Most SLP techniques focus at the routing level, with typically high message overhead. The objective of this paper is to investigate the novel problem of developing a TDMA MAC schedule that can provide SLP. We make a number of important contributions: (i) we develop a novel formalisation of a class of eavesdropping attackers and provide novel formalisations of SLP-aware data aggregation schedules (DAS), (ii) we present a decision procedure to verify whether a DAS schedule is SLP-aware, that returns a counterexample if the schedule is not, similar to model checking, and (iii) we develop a 3-stage distributed algorithm that transforms an initial DAS algorithm into a corresponding SLP-aware schedule against a specific class of eavesdroppers. Our simulation results show that the resulting SLP-aware DAS protocol reduces the capture ratio by 50% at the expense of negligable message overhead.