针对无线传感器网络中存在的安全问题, 提出了基于 Q-Learning 的分簇无线传感网信任管理机制(Q-learning based trust management mechanism for clustered wireless sensor networks, QLTMM-CWSN). 该机制主要考虑通信信任、数据信任和能量信任 3 个方面. 在网络运行过程中, 基于节点的通信行为、数据分布和能量消耗, 使用 Q-Learning 算法更新节点信任值, 并选择簇内信任值最高的节点作为可信簇头节点. 当簇中主簇头节点的信任值低于阈值时, 可信簇头节点代替主簇头节点管理簇内成员节点, 维护正常的数据传输. 研究结果表明, QLTMM-CWSN 机制能有效抵御通信攻击、伪造本地数据攻击、能量攻击和混合攻击.

This paper proposes a Q-Learning based trust management mechanism for clustered wireless sensor networks (QLTMM-CWSN) as a solution to the security problems in wireless sensor networks. This mechanism mainly considers three aspects: communication trust, data trust and energy trust. During network operation, the Q-Learning algorithm is used to update the node trust value based on the nodes’ communication behavior, nodes’ data distribution, and nodes’energy consumption. The node with the highest trust value in the cluster is selected as the trusted cluster head node. When the trust value of the main cluster head node in the cluster falls below the threshold, the trusted cluster head node replaces the main cluster head node to manage the member nodes in the cluster and maintain normal data transmission. The simulation results show that the proposed QLTMM-CWSN mechanism can effectively protect against communication attacks, forged local data attacks, energy attacks, and hybrid attacks.