Research On Spectrum Allocation And Decision Engine Algorithm For Cognitive Radios

Spectrum Allocation and Cognitive Decision Engine are important technologies incognitive radio. The goal of spectrum allocation is getting a series of idle spectrum afterspectrum sensing and allocating the idle spectrum to each cognitive user based on someallocation purpose in order to improve system utility. Cognitive decision engine adjusts theworking parameters based on the spectrum of wireless transmitted characteristic whencognitive users get idle spectrum, the engine can improve communication quality withmultiple objects.In terms of the problems of low convergence precision and complicated to achieve inspectrum allocation algorithm based on graph coloring model, the paper presents the spectrumallocation algorithm based on the differential evolution algorithm. Using the algorithm ofgood ability of finding the best and easy to achieve, we can improve corresponding utilityunder the allocation rules of maximize the system utility and proportional fairness.Meanwhile We can get better spectrum allocation scheme when the number of cognitive userand spectrum are large.The experimental results show using the algorithm presented in this paper can distinctlyimprove convergence precision and shorten running time, the scheme can perform better inmaximize system utility and proportional fairness.After getting allocation scheme, the working parameters of user can influence thecommunication quality. In terms of low convergence precision and poor in real-time, thepaper presents the cognitive decision engine based on the differential evolution algorithm.Using less parameters, easy to handle and achieve and good ability of finding the best, we canget better working parameters which can improve the efficiency of the decision engine.The experimental results show using the algorithm presented in this paper can get theworking parameters which distinctly improve system and different working goal performance.The running time is shorter than the best method of Co-evolutionary Particle SwarmOptimization algorithm.