Adaptive Trust Management Protocol Based on Fault Detection for Wireless Sensor Networks

Trust management is an important issue in selfconfigurable and autonomous networks such as Wireless Sensor Networks. Sensor nodes need to determine if other nodes are trustworthy, in order to decide whether to cooperate with them in completing the sensing and communication tasks. Therefore, adaptive trust management assures an appropriate level of security to the critical services provided by Wireless Sensor Networks. In this paper, we present the Adaptive Trust Management Protocol for Wireless Sensor Networks, which is able to compute data trust based on fault detection. The adaptive trust management protocol operates cycles in which reputation values are computed and penalty values are exchanged periodically. A spanning tree is generated for the sensor network, after which nodes evaluate their children using the fault detection mechanism and then exchange penalty values with their neighbors. The protocol has been implemented in TinyOS and evaluated in a test scenario using TOSSIM. Keywords-reputation, trust, fault detection, wireless sensor