网络节点重要度置信热评估模型研究与仿真

Research and Simulation on Thermal Evaluation Model of Network Node Important Degree Confidence

研究网络节点重要度评估优化问题。网络节点的重要度受节点邻近度与节点关键度的影响较大,一旦网络规模较大,局部节点的特性呈现较大差异,不同就局部节点的特性不同,很难形成统一判断,传统的评估方法只能从局部网络考虑网络节点间的关键度与邻近度,得出的网络节点重要度在局部有效,从全局来看,很难形成全局描述,给网络节点重要度的评估应用带来了困难。构建了一种网络节点重要度的置信热评估模型。基于模糊测度理论与Shapley熵对分布式节点之间的重要度置信进行动态热评估,通过模糊聚类的信息熵作为节点重要度置信评估的主要参考指标。对可测空间进行离散分解,得到模糊测度空间集合;对节点的信任度动态加载到信度函数中,利用Dempster规则对节点之间的重要度置信进行合成;以Shapley熵作为模糊测度的参考值并给出重要度置信分布。通过实验,利用改进方法在长期不诚实、长期诚实和偶尔不诚实三种特点的分布式节点进行分析,得到了较为理想的检测结果,并与其它重要度置信检测模型进行对比,成功率明显优于其它算法。

The evaluation optimization problem of network node importance degree was researched. The network node important degree is influenced by the node proximity and key degree greatly. Once the network scale is large, the local nodes show great differences, the characteristics of local nodes are different, and it is difficult to form a unified judgment. In the paper, a confidence thermal evaluation model of network node important degree was estab- lished, and on the basis of fuzzy measure theory and Shapley entropy, the important degree confidence among distributed nodes were processed with dynamic thermal evaluation. The information entropy of fuzzy clustering was taken as the main reference index of nodes important degree confidence evaluation. The measurable space was processed with discrete decomposition, and the fuzzy measure space set was obtained. The trust degree was loaded in the reliability function dynamically. The Dempster rule was used to synthesize the importance degree confidence value. The Shapley entropy was used as the reference value of fuzzy measure, and the confidence distribution was obtained. Through the experiments, the distributed nodes of three characteristics such as long - term dishonest, long - term integrity and occasionally dishonest were analyzed. Simulation result shows that the detection result is perfect, and the success rate is obviously better than other algorithms.