An Improved Method for Probabilistic Voting-based Filtering using Blacklists in Sensor Networks

Abstract

False report injection attacks and false vote injection attacks can be perpetrated easily by malicious attackers on the application layer in a wireless sensor network. These attacks drain the lifetime of the sensor nodes and prevent the forwarding of legitimate reports in the sensor network. A probabilistic voting-based filtering scheme (PVFS) was proposed in order to drop these two types of attacks simultaneously in intermediate cluster heads. Before transmitting a report, the scheme selects verification nodes within the intermediate cluster nodes to detect false votes attached from compromised nodes. In this paper, we propose a method to improve the detection power and energy savings by using a blacklist in every cluster head. The blacklist stores each compromised node ID and false key. The performance of the proposed method against these attacks was evaluated and compared to that of PVFS. The simulation results reveal that the proposed method enhances the average energy consumption and security level of each cluster head as compared with PVFS.