Research On Power Allocation Algorithm Of Cognitive Radio Based On Particle Swarm Optimization

The global wireless communication industry is booming,but it has to face a major bottleneck: how to allocate and utilize the wireless spectrum resources fairly and reasonably? Wireless spectrum resources are a very valuable non-renewable resource.For a long time,spectrum resources have been managed by static allocation.As a result,many wireless spectrum resources are often idle.Due to the issue of spectrum usage rights,other unauthorized users cannot use these idle wireless spectrum resources,resulting in a great waste of wireless spectrum resources.As a result,cognitive radio(CR)technology has received more and more attention.Its idea is to reuse the idle spectral resources without affecting the existing communication system.CR technology divides users into two types,primary users(PUs)and secondary users(SUs).The underlying spectrum sharing mode(Underlay)allows SUs to occupy the same spectrum resources as PUs at the same time,but the interference caused by SUs to PUs must be less than the interference threshold that PUs can withstand.This study focuses on how to allocate spectrum resources more efficiently in Underlay mode.As a new numerical search algorithm,intelligent optimization algorithm becomes an important way to optimize problem solving.Particle swarm optimization(PSO)has become a hotspot for intelligent optimization algorithms due to its versatility and fast convergence.Based on PSO algorithm,the following research on cognitive radio power allocation is made in this paper:For fairness among SUs,traditional power allocation systems often require strict Proportional fair among users.Such power allocation algorithms are complex,can result in higher calculation latencies that affect the quality of service(Qo S)of users,and strict proportional fairness results in lower overall system capacity.This paper presents an optimization model with channel capacity variance among SUs,transmission power of SUs and interference threshold of PUs as constraints.The model is simple,produces less computing time when power allocation is made,and uses variance constraints to set power for a single user more flexible,and increases system capacity.Because the model is a nondifferentiable,non-convex optimization power allocation problem,using ordinary PSO algorithm to solve,there are problems such as easy to fall into local optimization,slow convergence,and so on.This paper presents a PSO algorithm based on hybridization,which draws on the idea of hybridization in genetic algorithm and has obvious effects on improving the convergence speed and global search ability of the algorithm.The simulation results show that the algorithm not only achieves fairness among SUs and maximizes channel capacity,but also improves convergence speed and reduces algorithm complexity.Considering that many traditional algorithms assume that CR systems are in ideal channel states,such CR systems are not robust and cannot be implemented in practice.In order to solve this problem and continue to provide fairness to SUs in CR systems and to meet the Qo S needs of each user,a robust power allocation algorithm for cognitive radio networks based on hybrid PSO is proposed.In order to solve these complex optimization problems,adaptive weighting method and natural selection algorithm are combined.Adaptive weight method is effective in improving convergence accuracy by adjusting inertial weight according to the dynamic value of adaptability.At the end of each iteration,the natural selection algorithm ranks the particles according to their fitness value and replaces the poor ones with the good ones.The combination of the two algorithms can bring into play their respective advantages and improve the convergence speed and accuracy of the algorithm.The performance of the hybrid PSO algorithm in various aspects was verified through simulation experiments,and the results showed that the algorithm not only achieved the robustness and fairness of the system,but also had fast convergence speed and high convergence accuracy.The inertial weight of PSO algorithm can affect the algorithm’s optimization ability.The larger the inertial weight,the stronger the global search ability of PSO algorithm,but the local search ability will be worse.The smaller the inertial weight,the stronger the local search ability,but the global search ability will be poor.If the inertial weight can be adjusted according to the search characteristics of PSO algorithm,the PSO algorithm has better local optimization in the early iteration.It has faster convergence speed in the middle of iteration.By the end of the iteration,it has better local search capability.This PSO algorithm has good convergence speed and precision.Therefore,a new PSO algorithm based on sinusoidal adaptive weights method(SAW-PSO)is proposed,which makes the inertial weights of PSO algorithms change in sine function throughout the iteration,and can be adjusted according to the fitness value to make the inertial weights more suitable for PSO algorithm needs.In order to optimize the Qo S requirements of PUs and SUs,and to maximize the system capacity under certain conditions of fairness,constraints such as PUs transmission rate loss constraint,SINR constraint of SUs and launch power are added.Simulation results show that the power distribution scheme works well,and the algorithm has good convergence accuracy and convergence speed.