Improvement of energy consumption and detection power for PVFS in wireless sensor networks

In wireless sensor networks, adversaries can harm the sensor nodes by launching attacks such as the false report injection and the false vote injection. These attacks drain finite energy resources and drop legitimate event information in the sensor network. A probabilistic voting-based filtering scheme (PVFS) detects fabricated votes at the verification cluster-heads (CHs) while forwarding event reports. This paper presents a genetic algorithm-based PVFS to select effective verification CHs before transmitting the reports from a source CH. The proposed method determines the verification CHs based on the remaining energy level, the number of filtered votes, and the hop counts. The effectiveness of the proposed method is evaluated against the original PVFS in case of simultaneous multiple attacks. Our experimentation results indicated that the method results in an energy saving of approximately 10% and improves security against false report injection attacks and false vote injection attacks by 7% and 20%, respectively.