基于相依网络的无人机集群通信系统鲁棒性分析

Robustness Analysis of UAV Swarm Communication System Based on Interdependent Network

作为地震灾害救援和侦察战场等的主力设备,无人机集群通信系统的鲁棒性是抗毁性的重要指标,本文借助复杂相依网络理论建模无人机集群通信系统,引入相对网络效能比作为通信系统网络整体鲁棒性的指标,准确地反映节点之间的相互依赖关系,为评估通信系统在目标攻击下的鲁棒性提供了一种全新的视角。基于经典M-L模型和动态信息的负载重分配策略,建立相依同构ER-ER和异构ER-BA网络模型,考虑两层网络中相依节点度数的特性进行目标攻击,分析其在攻击下的级联失效动态过程,以及任意两节点的连边概率、平均度数和容量系数等特征参数,对通信网络整体鲁棒性的影响。研究表明,网络鲁棒性和网络相依强度负相关,在级联失效过程中可考虑降低52.09%~72.9%相依强度,以抵抗更严重的通信网络崩塌;对同构网络来说,相依模式的不同,RL、AL较DL整体鲁棒性分别降低了9.25%和15.95%。

As the main equipment for earthquake disaster rescue and reconnaissance battlefield,the robustness of UAV swarm communication system is an important index of destruction resistance.In this paper,we model the UAV swarm communication system with the help of complex dependency network theory,introduce the relative network effectiveness ratio as an index of the overall robustness of the communication system network,which accurately reflects the interdependence between nodes,and provides a new perspective for evaluating the robustness of the communication system under the target attack.A new perspective is provided for assessing the robustness of communication systems under target attack.Based on the classical M-L model and the load redistribution strategy with dynamic information,we establish the dependent homogeneous ER-ER and heterogeneous ER-BA network models,consider the characteristics of the degree of dependent nodes in the two-layer network for target attack,and analyze the dynamic process of their cascading failure under the attack,as well as the influence of the characteristic parameters,such as the probability of connecting edges of any two nodes,the average degree and the capacity coefficient,on the overall robustness of the communication network.It is shown that network robustness and network dependency strength are negatively correlated,and a reduction of 52.09% to 72.9% dependency strength can be considered in the cascade failure process to resist more serious communication network collapse.For isomorphic networks,the difference in dependency patterns has a greater impact by reducing the overall robustness of RL and AL compared to DL by 9.25% and 15.95%,respectively.