Multi-sensor Data Fusion Based on Consistency Test and Sliding Window Variance Weighted Algorithm in Sensor Networks

Jian Shu, Ming Hong, Wei Zheng, Li-Min Sun, Xu Ge

In order to solve the problem that the accuracy of sensor data is reducing due to zero offset and the stability is decreasing in wireless sensor networks, a novel algorithm is proposed based on consistency test and sliding-windowed variance weighted. The internal noise is considered to be the main factor of the problem in this paper. And we can use consistency test method to diagnose whether the mean of sensor data is offset. So the abnormal data is amended or removed. Then, the result of fused data can be calculated by using sliding window variance weighted algorithm according to normal and amended data. Simulation results show that the misdiagnosis rate of the abnormal data can be reduced to 3% by using improved consistency test with the threshold set to [0.05, 0.15], so the abnormal sensor data can be diagnosed more accurately and the stability can be increased. The accuracy of the fused data can be improved effectively when the window length is set to 2. Under the condition that the abnormal sensor data has been amended or removed, the proposed algorithm has better performances on precision compared with other existing algorithms.