Multi-population coevolutionary dynamic multi-objective particle swarm optimization algorithm for power control based on improved crowding distance archive management in CRNs

This paper aims to resolve the problem of power control in underlay CRNs better. Firstly, a multi-objective optimization problem of maximizing the throughput of PUs and SUs is proposed, which satisfied the constraints of PU’ interference temperature, the normal communication quality of all users and the transmission power limitation of users. Moreover, according to the theory of penalty function and particle swarm optimization (PSO), an improved multi-objective particle swarm optimization (IMOPSO) algorithm is proposed based on the archives management, which can achieve the maximum throughput of PUs and SUs as well as the minimum penalty constraints term. On this basis, in order to improve the boundary searching ability and diversity of the power control scheme, an improved coevolutionary multi-objective particle swarm optimization (ICMOPSO) in multiple population is proposed based on crowding distance archival management. Further, in order to adapt to the dynamic communication environment well, three different dynamic response schemes are presented correspondingly to cope with the instability of three types of environment. In the end, simulation results show that ICMOPSO and IMOPSO algorithm based on the archives management can obtain the maximum throughput compared with the conventional PSO. Through comparing with the performances in IMOPSO and ICMOPSO, it can be concluded that ICMOPSO algorithm has good abilities of stability, diversity and local search ability, which can provide more throughput optimal allocation schemes for decision makers and ensure the quality of customer service. On the basis of ICMOPSO algorithm, dynamic response strategy is better than the static response strategy at computational cost compared with the average number of iterations. And it can cope with the dimensional change of decision space in dynamic communication environment.