Dynamic channel allocation algorithm based on parameter estimation in cognitive radio

Dynamic channel allocation is studied for cognitive radio. Based on the characteristic of primary users´ service in cognitive radio, this article proposed that apply mixed Poisson distribution to build model for primary users´ service then estimate the probability density function of channel unoccupied duration by using EM algorithm to evaluate the parameter of mixed Poisson distribution model and addressed a multiple channel allocation algorithm based on the result of channel estimation. Numerical simulations have demonstrated the anticipated performance of the algorithm which proposed in the article at conflict rate and throughput. Meanwhile, the proposed algorithm have lower implementation complexity for practical reality.