Abstract: Aiming at the selective forwarding attack behavior of the wireless sensor network, a selective forwarding attack detection method combined with node characteristics and non-cooperative game (NC-NCG) is proposed. This method compares the forwarding rate in node characteristics with the threshold by setting up an independent monitoring network environment. The deviation degree between the current forwarding rate of the node less than the threshold node and the average forwarding rate in T time is calculated, and a secondary judgment is made according to the deviation degree to improve the detection rate of selective forwarding attacks. At the same time, in order to improve the network throughput, a non-cooperative game model with incomplete information is constructed to force suspicious nodes to participate in network functions and achieve accurate rapid node identification. Simulation results show that the method is not only effective in identifying selective forwarding attacks, but also improves network throughput and extends network lifetime with limited resources.

Key words: wireless sensor network, node characteristics, selective forwarding attack, non-cooperative games

摘要： 针对无线传感器网络的选择性转发攻击行为，提出了一种结合节点特征和非合作博弈的选择性转发攻击检测（node characteristics and non-cooperative game for selective forwarding attack detection，NC-NCG）方法。该方法通过设置独立监督网络环境，将节点特征中的转发率与门限阈值进行比较，计算小于阈值节点的当前转发率与T时间内平均转发率的偏离程度，根据偏离程度进行二次判定，以提高选择性转发攻击的检测率。同时为提高网络吞吐量，构建了不完全信息的非合作博弈模型，迫使可疑节点参与网络功能，实现节点快速识别。仿真实验结果表明，该方法不仅能够有效识别选择性转发攻击，而且在资源有限的情况下，可以提高网络吞吐量并延长网络生命周期。

关键词: 无线传感器网络, 节点特征, 选择性转发攻击, 非合作博弈