Research On Power Distribution Of Cognitive Radio Based On Particle Swarm Optimization

With the development of wireless communication technology,the 5G era is coming.Wireless communication has been closely related to people's daily lives.The increasing demand and high demand for service quality have made the radio spectrum scarcity problem increasingly obvious.The birth of cognitive radio brings hope to solve and alleviate this dilemma.As a smart wireless communication technology,cognitive radio realizes efficient use of spectrum resources through dynamic spectrum allocation and spectrum sharing.There are two users with different priorities in the cognitive radio(CR)network: one is called a primary user(PU),which has a high priority to have absolute use rights to the licensed band;the other is called a secondary user(SU),which has a relatively low priority and can only share the spectrum with the primary user in an 'opportunistic'manner while ensuring that the primary user does not interfere with normal operation.Power control is the key to realize spectrum sharing in cognitive radio network.Power control technology optimizes the transmit power of cognitive users by adopting different models and theories and adaptively adjusts the transmit power to ensure the quality of service(QoS)in cognitive radio network while achieving spectrum sharing.With the rise of intelligent optimization algorithms,more and more intelligent optimization algorithms and related improved algorithms are applied to the study of cognitive radio power allocation problems.Particle swarm optimization(PSO)has become a hot topic in this field.Based on particle swarm optimization and attempted to improve,this paper studies the power allocation problem in cognitive radio network for different constraints.Considering the interference temperature(IT)limit of the PU in the CR network,the data transmission rate of the SU,and the Signal-to-Interference-Noise-Ratio(SINR)requirement,Aiming at the problem that the traditional memory form of PSO can cause the algorithm to fall into local optimum,a novel particle swarm optimization algorithm based on new evaporation factor(LTPSO)is proposed to study the power distribution of CR network.The evaporation factor is set according to the particle swarm learning factor combined with the memory curve,and a new particle swarm memory form is established.The fitness value is scaled.The simulation results show that the LTPSO algorithm has a good optimization effect.In order to improve the transmission performance of the CR network,it is premisedon satisfying the IT constraints that the PU can tolerate,while taking into account the QoS of the SU and maximizing the capacity of the CR network system.The isothermal phase in the physical annealing process is introduced into the iterative optimization process of PSO.This paper proposes simulated annealing particle swarm optimization(SAPSO).The Metropolis criterion is used as the global optimal decision rule,and a dynamic inertia weighting factor is used to improve the fit of the annealing process.The simulation results show that the improved algorithm has better performance in both optimization and convergence.Based on the above parts,further considering the data transmission limitation of SU,in order to improve the optimization precision and expand the search space of particles,an adaptive simulated annealing particle swarm optimization(ASAPSO)algorithm is proposed.Adaptive control is used to dynamically adjust the particle swarm parameters according to the change of fitness value.The rules of the new solution are generated by Metropolis criterion,and the power distribution problem of CR network is further studied.The simulation results show that the ASAPSO algorithm has achieved good optimization results in all aspects.In the case where the constraints are relatively simple,that is,only the IT constraints of the PU and the SINR constraints of the SU are considered.This paper proposes a PSO based on genetic idea,namely genetic particle swarm optimization(GPSO),to study the problem of minimizing the cognitive power of cognitive users.In the PSO fitness value calculation,speed update and position update phase,the selection,crossover and mutation operations are respectively performed,and the corresponding secondary improvement algorithms are obtained and compared.Finally,all operations are introduced to obtain the GPSO algorithm.The simulation results show that compared with the Lagrangian multiplier method and PSO,the GPSO algorithm reduces the transmission power and obtains a higher SINR.