Research On OFDM System Resource Distribution Intelligent Algorithms

High speed, high bandwidth, high frequency spectra utilization efficiency andoffering various new services is the goal of wireless mobile communication developedmainly in future. There still have shortcomings such as small systematic capacity, lowutilization efficiency of frequency spectra and improper for transmiting high speed dataand multi-media service in the existing communication networks. Therefore, how tosupport stabile high speed data and multi-media business transmission reliably to be avery important program worthy of study under the condition of limited power andresource of frequency spectra etc..Orthogonal Frequency Division Multiplexing, a kind of carrier modulationtechnology which has good ability to deal with symbols and multipath interference andthe advantage of higher utilization efficiency of frequency spectra, have been appliedextensively in the fourth generation system of mobile communication, broadbandwireless access system as well as wireless city region net.In multi-user OFDM system,Self-adaption distribution algorithms of subcarriers and power which make use ofmultiuser diversity and frequency diversity could reduce error bit rate and power andimprove the frequency spectra utilization efficiency. At present there have been somealgorithms of subcarriers and power combined in some literatures, but it’s hard to adoptin practice because of the high complexity of these algorithms. Therefore the researchmethods of Rate Adaption mostly adopt two steps: subcarrier allocation and powerdistribution. The principle of subcarrier allocation is for each user to use the subcarrierswith high channel-to-noise ratio as much as possible while maintaining coarseproportional fairness. The objective of power distribution is to keep the proportionalfairness strictly on the basis of subcarrier allocation. Taking into account user fairnessand system throughput, Proportional rate constrains allocation algorithms achieved acertain degree of compromise between the two and make the system to achieve optimalperformance as a whole.The OFDM systems should support higher information rate and more flexiblequality of services because of insufficient wireless resource. As for resource allocationof OFDM system, the spectrum efficiency of the traditional algorithms is rather low.Intelligent optimization algorithm, a global optimization probability search algorithm which is highly parallelled、radom、adoption for itself, is developed by natural selectionand evolution mechanism of the biological field.They get applied widely in extensivefield because of the high efficiency optimization performance.The Ant Colonyalgorithm and Genetic Algorithm have representative.With the condition of equal powerassignment for all subcarriers, Ant colony algorithm is used in solving multiuser OFDMsystem subcarrier allocation with proportional rate constrains in the paper. Thealgorithm maximizes the sum capacity of multiuser orthogonal frequency divisionmultiplexing systems while maintaining proportional fairness. Due to the uniqueheuristic searching mechanism of Ant Colony algorithm, spectrum efficiency isguaranteed to converge quickly to an approximately optional solution. As for the slowconvergence speed and premature convergence problem of standard GeneticAlgorithm in solving power allocation of multiuser OFDM with proportional rateconstrains, this paper propose a power allocation algorithm of multiple populationgenetic strategy. The proposed algorithm use service fairness index as fitness value andthe retain generations of the best individual as the end rule. All population use differentcontrol parameter and contract with each other trough immigrant.Simulation resultsshow that the proposed algorithm’s convergence speed (almost100generation) is fasterthan standard genetic algorithm (almost300generation) and the result is steadier(almosttend to be zero). It can maintain the fairness of proportional rate while maximizing thesum capacity.