Intelligence Optimization Algorithm And Its Applications For Resource Allocation In Wireless Communication

With the rapid development of the times, people have a higher demand for the rate ofmultimedia services such as voice, data, and video. Therefore, how to use new wirelesscommunication technique for the demand of higher capacity becomes a hot research field. Wherein,OFDMA that can support mutipile user dynamic access is a typical key technology with highspectral efficiency, restraining multipath interference and flexible resource allocation. The dynamicOFDMA resource allocation will be the effective way to improve the entire system capacity andguarantee the quality of service in the future.As dynamic resource allocation of wireless communication belongs to schedulingoptimization problems, it will play a vital role for improving the performance of the entirecommunication system on how to allocate radio resource to users reasonably. Thus, thecorresponding resource allocation algorithms to solve such scheduling problems would haveemerged, which is mainly according to the method of convex optimization, distributed manner orheuristic. Combined with the current adaptive resource allocation algorithms, this paper primarilyapplies fish swarm algorithm, genetic simulated annealing intelligent algorithm and traditional dualoptimization algorithm to resource allocation, where both complete and imcomplete channel stateinformation will be considered. The algorithms furtherly improve OFDMA system performancesuch as system capacity, power and spectrum efficiency. Related research work as follows:(1) Aimming at OFDMA system with perfect channel state information, the rate maximizationproblem is studied and the resource allocation based on fish swarm algorithm is proposed. Firstly,the optimal subcarrier allocation is performaned. Then, power distrution will be done by the fishswarm algorithm with strong global optimization ability and easy realization. In order to ensure thefairness among users, a fitness function is given, which can measure the fairness and systemcapacity effectively. Finally, the simulation results indicate that the presented algorithm can obtaina better performance.(2) According to OFDMA system with perfect channel state information, two OFDMA adaptiveresource allocation schemes are presented. One is based on the proposed the criteria of energyefficiency. For the purpose of minimizing the overall power, the initial allocated bits andsubcarriers are adjusted optimally based on the criteria. The simulation results demonstrate that theproposed method of the scheme has a lower complexity and can minimize the total power,compared with the method of FDMA and classical distributed manner. The other is based on gentic simulated annealing algorithm. As a result of the addition of simulated annealing mechanism andinitial optimal individual, the algorithm can converge fast to achieve the minimum power. Thesimulation results can effectively verify the performance of the algorithm in the scheme.(3) In view of OFDMA system with imcomplete channel state information, the ergodic ratemaximization problem is studied. As the channel state information is not fully feedback to the basein the actual system, an OFDMA resource allocation scheme based on dual decomposition in theincomplete channel state information is proposed for obtaining the largest ergodic capacity. Here,In consideration of the case, channel state information is provided by the channel estimation valueplus a disturbance or error condition. In order to reduce the complexity of the algorithm, thescheme first relax the channel allocation factor to be a domain sharing factor, so that the originalproblem can be transformed into addressing a master-problem and sub-problem in thedecomposition model. Then, the optimal solution and approximate optimal Lagrangian value in themulti-user water-filling algorithm will be searched according to the iteration algorithm. In thealgorithm both variable-step size and fixed-step size are used. Finally, combined with requirementsof user’s quality of service, the capacity is maximized according to the approximate optimalLagrangian value. Theoretical analysis and simulation results show that the proposed scheme notonly can effectively improve the efficiency of resource allocation algorithm, but also enture thefairness of users while achieving the ergodic capacity maximization of resource allocation.