The Research Of Spectrum Allocation Technology In Cognitive Radio Network

Today,wireless communication technology has a rapid development,the spectrumresources are used more and more frequently, Actually, most of the spectrum that the primaryuser occupy is used unreasonable, these all cause the spectrum resources in shortage,How tosolve this question becomes one of the popular topics. In view of this situation, the concept ofcognitive radio is put forward.The spectrum allocation is a key teachnology,It is in response tothe detected wireless environment spectrum resources, make reasonable use of the spectrumresources that is idle temporarily.This paper mainly studies the spectrum allocation, The fulltext is divided into six parts, the specific work and innovation are as follows:Firstly, this paper describes the research background and significance of the cognitiveradio、main function and research status at home and abroad of spectrum allocationtechnology; then introduces its classification, principle and key technology andconstraints,introduces its several allocation models in detail.Based on this,we introduceseveral classical spectrum allocation algorithms based on gragh coloring theorymodel,through the simulation analysis and comparison of the distributed greedy algorithm anddistributed fairness algorithm we know: with the same spectrum number,the distributedgreedy algorithm can obtain greater network benefits than distributed fairness algorithm,andnetwork benefit decreases gradually with the increase of the number of users. However, thegreedy algorithm has a larger variance. through the simulation analysis and comparison of thecolor sensitive graph coloring algorithm and parallel spectrum allocation algorithm weknow:they obtain almost the same network benefit with the change of spectrum number. Butthe time overhead of csgc almost linearly increases with the increasing of spectrum numberand the time overhead of parallel algorithm basically unchange,it shows a linear state.Finally, this paper mainly study the spectrum allocation algorithms based on geneticalgorithm,the quantum genetic algorithm, then compare them with csgc.the genetic algorithmand the quantum genetic algorithm have many iterations, easy to fall into local optimumproblem, in order to solve these problems this paper proposes an improved quantum geneticalgorithm, it use the new quantum rotation gate strategy updates the population, takecatastrophe strategy when the population fall into local optimum.The improved algorithmensure the diversity of the population and avoid the local optimum effectively,also,it has afaster convergence rate relative to the csgc and quantum genetic algorithm,can improve network benefit better. Then on the basis of the spectrum allocation study based on particleswarm optimization algorithm, simple quantum particle swarm optimizationare presented forthe discrete optimization problem, it has stronger ability of searching global optimal solution,avoid the particle swarm algorithm may fall into a local optimal solution phase, and comparewith the pso、csgc,it can get better network benefit.