【摘要】 光纤网络具备动态变化的特性，使得其节点与连接可能随时出现增减情况。由于光纤网络中存在大量的节点和连接信息，在监测过程中，很可能会出现误将健康节点判定为入侵点，或是遗漏真正入侵点的状况，增加了准确区分健康节点与入侵点的难度。为了精准筛选出光纤网络中的健康通信节点，文章针对入侵环境提出了一种基于结构洞和K核重要性的健康通信节点筛选方法。采用种群聚类改进的粒子群优化（particle swarm optimization,PSO）算法进行光纤网络入侵检测。通过聚类操作，将相似的粒子聚集在一起形成多个子种群，实现独立搜索与信息交流，以避免早熟收敛问题。基于结构洞理论识别光纤网络入侵环境中的关键节点，并利用K核重要性确定相邻节点间的重要度贡献，评估节点的稳定性和影响力，从而筛选出其中的健康通信节点。实验结果表明，所提方法在不同的光纤网络入侵环境下，均可以筛选出健康通信节点，降低了错误筛选和漏筛的概率，可以更加有效且更加准确地筛选出健康通信节点。更多还原

【Abstract】 Fiber optic networks have the characteristic of dynamic changes,which means that their nodes and connections may increase or decrease at any time.Due to the large number of nodes and connection information in fiber optic networks,it is possible to mistakenly identify healthy nodes as intrusion points or miss the true intrusion points during monitoring,which increases the difficulty of accurately distinguishing healthy nodes from intrusion points.In order to accurately screen healthy communication nodes in fiber optic networks,this paper proposes a health communication node screening method based on structural holes and K-core importance for intrusion environments.Adopting the particle swarm optimization(PSO)algorithm improved by population clustering for fiber optic network intrusion detection.By clustering operations,similar particles are gathered together to form multiple sub-populations,enabling independent search and information exchange to avoid premature convergence problems.Based on the theory of structural holes,identify key nodes in the intrusion environment of fiber optic networks,and use K-core importance to determine the importance contribution between adjacent nodes,evaluate the stability and influence of nodes,and select healthy communication nodes among them.The experimental results show that the proposed method can screen for healthy communication nodes in different fiber optic network intrusion environments,reducing the probability of incorrect and missed screening,and can more effectively and accurately screen for healthy communication nodes.更多还原