Research On Topology Control And Intrusion Detection Of WSN Based On Game Theory

With the continuous development of sensor technology,wireless communication technology and microelectronics technology,wireless sensor network is increasingly popular.It is widely used in the fields of environment,industry,military and home furnishing.The network can sense and collect object information in the target area in real time,which improves the ability of human beings to understand the objective world.The limited resources of sensors have always restricted the development of wireless sensor networks,so how to reduce network energy consumption and ensure network security is the key research content of wireless sensor networks.Topology control and intrusion detection technologies can meet the energy efficiency and security requirements of wireless sensor networks to a certain extent.As a mathematical theory to studying competition phenomena,game theory provides a new research idea for wireless sensor network,which is applied in topology control and intrusion detection by many scholars.This thesis takes the energy saving and security requirements of wireless sensor networks as a starting point to carry out in-depth research on issues related to topology control and intrusion detection with game theory.Firstly,this thesis studies energy-efficient and fault-tolerant network topologies by using non-cooperative game theory.The energy of sensor is limited and not easy to recover,and the deployment environment is complex,which requires that the network topology has lower energy consumption and a certain fault tolerance.In response to these needs,based on the non-cooperative game theory,the relationship between network connectivity,power and energy is weighed in the design of the utility function,and a topology control game model was established.It is mathematically proved that there is at least one Nash equilibrium in the game.On this basis,an energy-efficient and fault-tolerant topology control game algorithm(EFTCG)is proposed.The algorithm satisfies the needs of energy saving and fault tolerance of topology control,realizes network topology adaptation.Under the condition of network single connectivity,it has good energy-saving effect and effectively prolongs the network lifetime;Under the condition of network dual connectivity,it has better fault tolerance performance and longer network life cycle.Secondly,this thesis uses evolutionary game theory to build a reasonable evolutionary game model of intrusion detection to solve the problem of balance between efficiency and energy consumption of intrusion detection system.The operation of intrusion detection system needs a certain amount of computing resources and energy overhead.However,the battery energy,computing power and storage capacity of the sensor itself are very limited.Therefore,in order to effectively use the intrusion detection system,we must consider the characteristics of limited network resources and choose appropriate strategies to configure the system.Based on the assumption of bounded rationality,this thesis analyzes the attack and defense process of intrusion detection using evolutionary game theory.At first the optimal response dynamic mechanism is used to analyze the attack behavior of malicious sensors.It can be concluded that the malicious sensor nodes always attack the cluster head,so the cluster head should always turn on the intrusion detection system for defense.On the basis of this analysis,this thesis uses the replication dynamic equation to analyze the evolution trend of attack and defense strategies under various attack modes,and proposes the optimal defense strategy selection algorithm of intrusion detection system.The intrusion detection system can actively and dynamically adjust the detection strategy for effective defense according to the trend of attack and defense evolution.On the premise of ensuring the detection rate,it reduces the energy consumption,balances the system performance and network energy consumption,and ensures the long-term security of the whole network.