Abstract: For the traditional spectrum decision methods in cognitive radio networks, the secondary users select their communication channel based on various criteria, which may result in channel contend and congest. Considering probability-based spectrum decision in cognitive wireless mesh networks,a greedy channel selection algorithm was proposed. When spectrum handoff occurres, the proposes algorithm combined with the improved preemptive resume priority (PRP) M/G/lqucuing theory to select optimal target channels for secondary users. Numerical analysis results show that the algorithm can increase and balance the channel utilization. With lower average arrival rate of secondary users,the average overall transmission time of the secondary users has been further reduced and the network performance isenhanced.

Key words: Cognitive wireless Mesh networks, Spectrum decision, Greedy algorithm