Grouping-based Resilient Statistical En-route Filtering for Sensor Networks

In sensor networks, the adversaries can inject false data reports from compromising nodes. Previous approaches for filtering false reports, notably statistical en-route filtering, adopt a simple strategy for grouping sensor nodes that requires redundant groups and thus decrease the filtering effectiveness. Worse still, they either suffer a threshold problem, which may lead to complete breakdown of the security protection when the threshold is exceeded, or are dependent on sink stationarity and specific routing protocols, which cannot work with mobile sinks and various routing protocols. In response to these, this paper proposes a scheme, referred to as Grouping-based Resilient Statistical En-route Filtering (GRSEF), in which nodes are grouped once deployed without requiring redundant groups and a location-aware approach based on terrain division along multiple axes is proposed for key derivation. The design of GRSEF, which is independent of sink stationarity and routing protocols, provides a well suitable en-routing filtering solution for sensor networks with mobile sinks. Analytical and simulation results verify that the scheme significantly improves the filtering effectiveness and efficiently achieves the resiliency against node compromise.