The Dynamic Spectrum Sharing Based On Auction Model In Cognitive Radio Networks

Cognit ive radio technology is one of the most effective methods to tackle the problem of unbalance use of spectrum resource.Dynamic spectrum allocation is a key step for achieving the spectrum resource sharing.According to the auction theory,this thesis does a research on the dynamic spectrum allocation and modif ies genet ic algorithm for a better plan of the spectrum allocation.The details are shown follows:(1)For spectrum allocation in cognitive networks,this thesis apply the mechanism of second-price sealed auction to solve the problem,which can guarantee an authentic network.The auction mechanism is enhanced and the generated allocat ion plan can support the existing spatial reuse of spectrum resources technology better to achieve resource sharing.Besides,the mechanism not only maximizes the primary users' profit but also enhances the secondary users' satisfaction.(2)For the spectrum auction problem in a cognit ive network with one primary user and mult iple secondary users,this thesis do a research and provide a solution called hybrid genet ic algorithm based on greedy strategy.This thesis also exploit a payment function for secondary users based on the second-price sealed auction mechanism and prove the authenticity of the network.The experimental results show that compare with the conventional greedy algor ithms and hybrid genetic algorithms,the improved hybrid genet ic algor ithm solve the problem of spectrum sharing much better in the cognitive network w ith one primary user.It helps to improve the primary user's welfare and enhanc e the second users' satisfaction.(3)For the spectrum auction problem in the cognitive network with multiple primary users and mult iple secondary users,this thesis provides a solution called Non-dominated Sorting Genet ic Algor ithm II(NSGA-II).The crowd mechanism of the algorithm is improved to avoid the loss of excellent genes and boost the efficiency of the algorithm.The experimental results show that compared to the traditional NSGA-II algorithm and the improved hybrid genet ic algorithm with weight coefficient,the improved NSGA-II algorithm can help primary users increase their earnings and helps secondary users enhance the satisfaction.This thesis focuses on the dynamic spectrum allocat ion problem in cognitive networks.Given that primary users release,the improved auction and algor ithms we propose can maximize both the primary users' earning and secondary users' satisfaction.In the allocation plan,several secondary users may gain the same spectrum resource and they are able to share the same resource by existing spatial reuse technology.The experimental results show that the improved auction mechanism and algor ithm can help primary users get more income and help more secondary users obtain the spectrum resource,also can guarantee an authentic network.