Research On Spectrum Management For Cognitive Radio Networks In Dynamic Spectrum Market

The exponentially growing demand for spectrum to satisfy the greed for more applications has come to a bottleneck state since the spectrum resources are allocated in a centralized, static and wholesale fashion. Cognitive radio networks have emerged as a promising technique to improve the utilization of the existing spectrum.A market-based spectrum management approach is expected for successful use of cognitive radio.We focus on dynamic spectrum allocation based on spectrum market. In section 3, we propose a highly available spectrum allocation service (HA-SAS) model based on replication in space, which finds the best set of channels to meet more secondary users'reliability requirements with minimum price. In this model, we use parallel mechanism to improve the transmission reliability to meet secondary users'requirements, and to meet the secondary users'reliability requirements with minimum cost, we propose an effective algorithm, which sorts the available channels by a function of price and reliability, and then selects target channels iteratively until secondary user's request can be satisfied. Finally, the job is replicated to the selected group. Simulation results show that spectrum utilization improved nearly 25% compared to No-HAS, and purchase price from secondary users is relatively stable with our proposed model.In section 4, we mainly focus on dynamic spectrum allocation based on spectrum market when requirements from secondary users are not urgent. McAfee and Trust spectrum allocation method is used in our problem model which we assume all secondary users can use the spectrum that offered by primary users in one area. Finally, we propose a highly available spectrum allocation service (HA-SAS) model based on replication in time.