Improving Data Reliability via Exploiting Redundancy in Sensor Networks

Abstract Wireless sensor network applications interact with the physical world through analog sensors. However, decisions derived from flawed sensor measurements can adversely impact the correctness of the overall sensor network findings. To improve the reliability of decisions and minimize the impact of faulty sensor measurements, it is important to have a distributed scheme that enhances the trustworthiness of result by exploiting the presence of redundant data. In this paper, we present Confidence Weighted Voting (CWV), a distributed technique that can improve the underlying data reliability and fault tolerance of various sensor network applications. We examined CWV against Majority Voting (MV) and Distance Weighted Voting (DWV) techniques, and contrasted the level of data reliability of each approach in the prevalent presence of flawed sensors. The results show that CWV can consistently and substantially outperform the other two distributed schemes by providing as much as 49% more resiliency against faulty sensor measurements.