Threat-aware clustering in wireless sensor networks

Technological advances in miniaturization and wireless networking have enabled the utilization of distributed wireless sensor networks (WSN) in many applications. WSNs often use clustering as a means of achieving scalable and efficient communications. Cluster head nodes are of increased importance in these network topologies because they are both communication and coordination hubs. Much of the research into maximizing WSN longevity and efficiency focuses on dynamically clustering the network according to the residual energy contained within each node. This is a result of the commonly held assumption that battery depletion is the primary cause of node failure. In this work, we consider that there are applications in which threats may significantly impact node survival. In order to cope with these applications, we present a threat-aware clustering algorithm, extending the Hybrid Energy Efficient Distributed clustering algorithm (HEED) that minimizes the exposure of cluster heads to threats in the network environment. Simulation results indicate that our extended threat-aware HEED, or t-HEED, improves both the longevity and energy efficiency of a WSN while incurring minimal additional overhead. Our research demonstrates and motivates the need for a general framework for adaptive context-aware clustering in WSNs.