Research On Resource Allocation Algorithm In OFDM Cognitive Radio System

The rapid growth of wireless communication technology and the rapid developmentof wireless communication services result in the rapid demand of the valuable wirelessspectrum resource. Cognitive radio (CR) as an intelligent spectrum sharing technology, canimprove the efficiency of the spectrum utilization by making the unauthorized usersopportunistically access the idle spectrum. CR system can perceive wirelesscommunication environment using the sensing mechanism, detect the idle spectrum, andchange working parameters of the system (such as power, modulation mode, transmissionrate) adaptively according to some decision-making algorithms to efficiently utilize thespectrum resource. Orthogonal frequency division multiplexing (OFDM) technology withflexible subcarriers configuration capability and the power control capability, the OFDMdynamic resource allocation can adjust subcarrier and power allocation based on theinstantaneous channel status information, which further improve the spectrum utilizationefficiency and system performance. In frequency selective wireless fading channel, thecombination of cognitive radio technology and OFDM dynamic resource allocationtechnology can give play to the advantages of them. Therefore, it has received more andmore extensive attentions and researches.Other important task of resource assignment to SU in the CR system is to detectavailable idle spectrums unused by PUs using sensing mechanism. To accurately sensethese spectrums is a very difficult task due to the multi-path loss, stochastic wirelesschannels, attenuation and the interference from SUs. Therefore, sensing result can not becompletely reliable; and always produce some detection errors because of the channelconditions between the transmitter of PU and the sensor of SU. The error detectionprobability will increase the interference to PUs, and the false alarm probability will limitthe throughput of CR system.Traditional resource allocation schemes have already obtained some good researchresults, and reached a higher total capacity under the total power constraint for SU and theinterference constraint of PU at the same time. However, the above resource allocationschemes are all based on perfect spectrum sensing. In practical CR systems, the effect ofimperfect spectrum sensing should be considered.In consideration of resource allocation problem for OFDM-based multi-user cognitiveradio systems with imperfect spectrum sensing, firstly, this paper has analyzes thespectrum sensing, and calculate the interference caused to PU from SU through theprobability of PUs in using subcarriers, the misdetection probability and the false alarm probability of the sensing system. Then take this interference as an optimization condition,and a resource allocation algorithm is proposed such that the capacity of the CR system ismaximized while keeping total transmission power and the interference to each PU withinprescribed thresholds and meeting the requirement of proportional rate for cognitive user.To reduce the computational complexity, the resource allocation is in two steps, subcarrierassignment and power allocation.Simulation results show that the proposed scheme can achieve lower interference toPUs when considering spectrum sensing error than that the situation without considerationof the error, meanwhile, system total capacity only reduce little and it will decrease withthe increase of misdetection probability. Compared with the dynamic resource allocationtechnique of the traditional multi-user OFDM system, the algorithm proposed in this paperis a valuable attempt to consider the resource allocation in a cognitive radio which is moreclose to the actual system.