Entropy-Based Trust Management for Data Collection in Wireless Sensor Networks

In wireless sensor networks, the sensor nodes are used to collect observed data. But they may be interfered by the environment or attacked by an adversary. If that, the accuracy and reliability of the data will be affected. To avoid this problem, we propose a lightweight trust management framework for the multi-level cluster sensor networks by adopting the entropy-based trust model. The area trust of the cluster head is calculated from the propagation trust and the area trust of its cluster members. When the area trust of a cluster head drops below a threshold, it means that there are some exceptions in the data transmission of its area. Then the cluster head will detect these exceptions and repair them locally. From the analysis of the memory and communication overhead, it shows that this framework is lightweight and practical.