Trust computation model based on improved Bayesian for wireless sensor networks

To realize the fast decline and slow growth of the trust value, the penalty factor and the reward factor are introduced to modify the Bayesian equation of trust model in the wireless sensor network (WSN). In addition, attenuation function is used to update the trust value. The confidence of direct trust is used to determine whether the direct trust is trustful enough to be the overall trust, and this can reduce energy consumption and the impact of malicious node. The simulation shows that the trust model is accurate and reliable, and can effectively save energy.