Abstract: Cognitive radio is regarded as a promising technology to improve spectrum utilization significantly. Studies are put to discuss underlay spectrum sharing and power control, but issues such as the interference of the primary system are also considered as the constraint. The article builds up a spectrum allocation mathematical model which considers different interference intensity according to relative geographic locations between two SLs in the spectrum-sharing mode of cognitive radio network. Then it is converted into multi-objective optimization problem. To solve the spectrum sharing problem, the multi-objective improved genetic algorithm was adopted. Simulation shows that the proposed methods greatly outperform the commonly used K-max-cut in graph theory. It can realize the network benefit maximization and reduce the disturbance to the primary system by using the multi-objective optimization algorithm.

Key words: cognitive radio network, spectrum sharing, spectrum allocation, improved genetic algorithm