Research of Cognitive Engine Technology Based on Improved Artificial Fish Swarm Algorithm
|School||Harbin Engineering University|
|Course||Communication and Information System|
|Keywords||Cognitive radio Decision engine Artificial Fish Swarm Optimization behavior Throughput optimization|
As the core technology of cognitive radio, the basic function of cognitive radio decisionengine is to intelligently optimize the allocation of radio transmission parameters based onuser requirements and the environmental change. Under the premise of authorized users’normal communication, maximize the use of spectrum resources in order to providecommunication access services for cognitive users and ensure the quality of communication.Finally achieve the purpose of improving spectrum utilization and reasonable optimize the useof various wireless communication resources.While cognitive decision engine parameter optimization problems can be summed up asmulti-objective optimization problems, also using intelligent algorithm to solve themulti-objective optimization problem is the current mainstream trend, so intelligent algorithmis widely used in cognitive decision engine. But when solving the problem of parameteroptimization by cognitive decision engine, the traditional algorithm has a slow convergencespeed, and easy to fall into local optimal solution and other defects, which has influenced theaccuracy and speed of the solution of the problem, so this paper introduced Artificial FishSwarm Algorithm (AFSA) to the research of cognitive radio decision engine.The core idea of Artificial Fish Swarm Algorithm is to use the process of fish schoolchasing food to imitate the optimization process, this paper systematically researchedArtificial Fish Swarm Algorithm (AFSA),since traditional Artificial Fish Swarm Algorithm(AFSA) has slow convergence speed, easy occur the extreme value point nearby cycleconcussion and other defects, and exploited three sets of improvement schemes, the core ideasare increasing optimization behavior, utilizing adaptive adjustment parameter to controlhorizons step length of foraging behavior and appropriately shielding random behavior. Whilein this paper, those three improvement schemes were carried on system function test andproved their advantages that compared with traditional Artificial Fish Swarm Algorithm(AFSA). Then fused and applied them into cognitive radio decision engine.The experiments show that in the multi-carrier communication system, this cognitiveradio decision engine has a high convergence precision, a high average fitness value andstrong stability and some other characteristics,its performance is significantly better than current advanced Binary Quantum Particle Swarm (BQPSO) cognitive engine.Finally, this paper has systematically analyzed the functional relationship betweenperception time and the effective throughput of cognitive users in fixed frame long energydetection model as well.And also the applied this improved Artificial Fish Swarm Algorithm(AFSA) into optimization problem of effective throughput of cognitive users. Utilizing theadvantages of fast convergence speed, high convergence precision, strong climbing ability,good global convergence, the improved Artificial Fish Swarm Algorithm (AFSA) has greatlyincreased scope of application and perceived accuracy of the algorithm, the convergence rateis significantly better than the traditional.