A Service-Oriented Spectrum Assignment Algorithm for Cognitive Wireless Networks

Cognitive Radio is a promised solution to conquer the inefficiency utilization of spectrum resources, which is induced by the traditional fixed spectrum allocation scheme. In cognitive wireless networks, deficient matching between users´ demands and spectrum resource may cause communication deterioration while superfluous matching may cause spectrum inefficient utilization. In this context, by considering multiple spectrums´ attributive characters and the cognitive users´ service type, a service-oriented adaptive spectrum allocation algorithm based on Graph-Coloring is proposed to improve the match degree between service demands and spectrum resource, so that more efficient spectrum utilization could be achieved. The simulation results demonstrate the proposed algorithm can better satisfy the user´s demand and achieve better utility reward and fairness than a set of existing algorithms.