考虑级联失效的有向WSNs节点重要度评估模型

Evaluation Model of Node Importance Considering Cascading Failures in Directed WSNs

针对无线传感器网络中关键节点失效引发网络快速崩溃的问题,考虑节点之间存在相互作用且相互作用具有方向性的特征,建立有向网络级联失效模型,推导节点失效后引起的负载震荡状态值.基于经典的PageRank算法,利用失效节点引发的邻居节点平均负载震荡状态值定义节点的初始重要度值,建立节点度择优的分配规则改进PageRank算法的平均分配规则,结合节点层级结构,建立考虑级联失效的有向传感器网络节点重要性分析方法.最后,通过仿真实验验证算法的有效性.实验证明,该方法能有效地评估了有向无线传感器网络节点的重要性,而且当对判定出的关键节点进行保护时,该方法判定关键节点的网络抗毁性明显较强.

To solve the problem of network collapse quickly caused by the critical nodes’failure in wireless sensor networks,the interaction between nodes and the directionality of interactions are considered and a cascading failure model of directed networks is proposed.The node load oscillation state values caused by the node after its failure are derived.Based on the classical PageRank algorithm,the average oscillation state value of neighbor nodes caused by cascading failures is taken as the initial importance value of nodes.The node degree preferential rule is built to improve the irrationality of the average allocation in PageRank algorithm.Then,a method of node importance evaluation method for directed sensor networks considering cascading failures is built.We verify the effectiveness of the proposed algorithm by simulation experiments.The experimental results showthat this algorithm can effectively evaluate the importance of directed wireless sensor network nodes.Moreover,protecting the key nodes determined by this method can obviously enhance the invulnerability of the network.