EVALUATION METHOD OF SENSOR DATA CREDIBILITY BASED ON MULTI-SOURCE HETEROGENEOUS INFORMATION FUSION

The reliability of sensor node data is crucial for data processing, especially sensor network data for security monitoring. However, existing evaluation methods cannot effectively evaluate the data credibility when the sensors are occluded or damaged. In order to solve this issue, a credibility assessment method for sensor data is proposed in this paper by fusing multi-source heterogeneous sensor data. Experimental results on mine smoke data show that the proposed method could effectively improve the veracity of evaluation when the sensors are disturbed or blocked.