Research On Ad-hoc Network Reputation Management Mechanism Based On Multi-agent System Consensus Theory

Ad-hoc network is a kind of centerless multi-hop network,and its network-ing and wireless communication at any time provides convenience for practical scenarios,and is widely used in military,civilian and commercial fields.Com-pared with ordinary wired network,Ad-hoc network is not high security.Ad-hoc network communication relies on cooperative forwarding between nodes,so judg-ing the reputation of a node and eliminating malicious nodes are very important for the safe operation of the Ad-hoc network.Reputation system is a widely used technology to identify and deal with non-cooperative nodes.The reputation system is introduced in the Ad-hoc network.Nodes are evaluated based on the reputation of other nodes directly observed and the reputation spread between re-ceiving nodes,which can effectively handle behavior Mischievous nodes improve network security,reliability and information.A good reputation management system has important strategic significance for the safe and stable operation of China's mobile Ad-hoc network system.The multi-agent system is composed of a large number of individuals(nodes)and their interactions(edges).The individuals in the system have certain inde-pendence and autonomy.All individuals have only local perception capabilities,and no individual has the overall system.information.The main idea of distribut-ed collaborative control in multi-agent systems is to design distributed control al-gorithms based on individual communication between individuals in the system,thereby driving individual evolution to achieve the desired global collaborative behavior.The characteristics of local communication,autonomy,parallelism,and coordination emerging from the group behavior of mobile Ad-hoc networks in a complex communication environment are consistent with the collaborative char-acteristics of multi-agent systems,making distributed cooperative control theory of multi-agent systems It provides a good mathematical model and theoretical basis for the reputation management of the mobile Ad-hoc network.Based on the above analysis,this study mainly considers mobile Ad-hoc net-works in complex communication environments.Based on the distributed conver-gence algorithm theory of multi-agent systems,a reputation management mech-anism is designed in mobile Ad-hoc networks.This article focuses on the following issues:To give a preliminary assessment of the reputation of other mobile nodes.In a complex communication environ-ment,can a mobile node use only local information dissemination to manage its reputation with other nodes.The main work contents are as follows:1)We considered a noisy communication environment that is closer to the actual communication environment.Based on a stochastic approximation conver-gence algorithm,we establish a reputation management mechanism for the mobile Ad-hoc network in a randomly generated topology graph,and analyze the impact of noise on the reputation management system.2)We also considered a two-valued communication environment.A two-valued communication environment can reduce the communication bandwidth re-quirements during the propagation of reputation information,thereby speed-ing up the propagation of information and avoiding channel crowding.When establishing a reputation management mechanism,we use the empirical mea-surement method to estimate the received binary information of neighbor n-odes,and then use the latest convergence algorithm with two-valued com-munication to establish a reputation management mechanism.At the same time,under different communication-control structure models,we also quan-titatively analyzed the influence of malicious nodes and other factors on the reputation management of mobile Ad-hoc network systems.The research results of this paper will provide solid theoretical support for the construction and operation of mobile Ad-hoc network reputation management,and it is also an effective attempt to apply the convergence theory to engineering practice.