Energy Saving Replay Attack Prevention in Clustered Wireless Sensor Networks

Abstract

Nodes in a wireless sensor network (WSN) are susceptible to various attacks primarily due to their nature of deployment and unguarded communication. Therefore, protecting such networks from attacks is a major challenge. We propose a scheme where sink, the resource-rich node, keeps track of attack threat based on statistics of network traffic. If any threat is apprehended, the sink and nodes together start analysis on each node’s contribution towards the network traffic to confirm the attack (if any) and identify the attacked node. Unlike existing attack prevention scheme, which incorporates securing mechanism starting from the network in operation, the present scheme starts secure communication only when it detects attack thereby reducing securing overheads significantly. As there is energy constraint in WSN, the main objective of this work is to reduce energy consumption while keeping the network secured. The other objective is to reduce attack detection time concomitantly so that the network can start secure communication before the attack causes any significant damage in the network. Performance of the scheme is compared with one of the state-of-the-art scheme. Qualitative analysis is done considering the performance metrics of communication, computation and storage overheads while quantitative analysis is done considering energy consumption, detection time and detection probability as performance metrics. Results confirm significant improvement over the competing scheme in respect to all the overheads.