高效混合式协同检测方法研究

针对单个节点检测性能容易受多径、阴影效应的影响,提出一种基于模糊聚类和D S证据理论的协同检测方法:首先依据协同节点的拓扑结构和距离信息,采用模糊C均值算法对节点聚类,然后利用OR准则对各聚类内的检测结果进行融合,各聚类之间采用D S证据理论进行融合;最后建立博弈概率分布,根据最大概率原则做出最终判决。仿真结果表明,在可控计算复杂度的情况下,该方法检测性能比传统方法有很大提高,而且在虚警概率为10%时,检测概率能达到90%以上。

Research on High-efficiency Mixed Cooperative Detecting Method

Performance detection based on single node is easily affected by muhi-path and shadow, so a cooperative detecting fusing method based on fuzzy clustering and D-S evidence theory is proposed. First, fuzzy c-means algorithm is used to realize cooperative nodes clustering. Second, OR-rule is used to fuse the detecting results of all cooperative nodes in the same cluster. Third, D-S evidence theory is used to fuse the fused results of all clusters in the same cooperative area. At last, probability distribution is established based on game theory and final decision is made based on the maximum probability principle. Simulation results show that the presented method with controlled calculation gives better detecting performance than the traditional ones. And the detection probability is above 90% when the false probability is 10%.