摘要

Addition Encouragement and Multiplication Punishment (AEMP) trust model is proposed for Wireless Sensor Networks (WSN) to block attacks from network. The core idea of the model is to calculate a trust value of each node according to its communication behavior. A good behavior results in a bigger value while a bad behavior leads to a lower value. The model consumes less resource and has higher performance compared with other existing models. The model has three levels: basic AEMP, AEMP model integrated with direct &indirect communication trust value and AEMP considered malicious evaluation. Simulation experiments have been done to test models'performance. AEMP model has a number of applications in WSN, such as finding out bad nodes in WSN or designing a routing protocol.