Mobile Agent Cooperative Sensing Algorithm Based On Swarm Intelligent In Cognitive Wireless Network

Due to the rapid development of wireless communication and the continuous increase of mobile devices,the shortage of spectrum resources is becoming increasingly prominent.Cognitive radio can dynamically access the idle spectrum without affecting the primary user.As a result,the utilization rate of spectrum has been improved and the rapid development has been achieved.Spectrum sensing is a prerequisite for the application of cognitive radio.However,with the increase of portable mobile devices,the mobility of users greatly increases the challenge of spectrum sensing,which is closely related to signal state,signal coverage and device mobility.It is not possible to provide effective perceptual data only by secondary users at fixed locations.Therefore,in dynamic cognitive radio networks,the collaboration of secondary users will effectively improve the accuracy of perceived data.In this thesis,we propose a mobile agent cooperative sensing algorithm based on swarm intelligence for the primary user mobility scenario.The main contents are as follows:(1)The problems of standard particle swarm optimization in dynamic environment are discussed,and a distributed particle swarm optimization algorithm for finding multi-objective is proposed.Each particle updates its speed and position independently,and only notifies other particles when better fitness value is found.Limits the maximum speed of the particle and avoids increasing the difficulty of searching because of jumping out of the search range.In addition,a correction mechanism has been added to prevent all particles from moving towards the same target and losing diversity.(2)In dynamic environment,secondary users do not have prior knowledge about the primary users and cannot obtain their moving tracks.Tracking the primary users will also affect the needs of the secondary users themselves.In this thesis,a mobile agent cooperative sensing algorithm based on multi-objective particle swarm is proposed to deploy mobile agents in the network.Distributed particle swarm is used to accomplish tasks through inter-agent cooperation,which reduces the cost of hardware input and energy consumption,and at the same time,avoids the situation in which secondary users participate in spectrum sensing but cannot be allocated to the spectrum.Each agent is mapped to a particle to determine the primary user signal based on the RSS.The agents that sense the same primary user constitute the neighborhood,and the distributed data fusion is used to obtain the convergence value,which is compared with the energy threshold to obtain the final judgment.Finally,the simulation scheme is designed.The experimental results show that the proposed method can solve the influence of the primary user movement on spectrum sensing,and verify the feasibility of the scheme.