一种新颖的低计算复杂度OFDMA资源分配算法

A Novel Low-Complexity Algorithm for Source Allocation of OFDMA System

在融合遗传算法和解析方法的基础上,提出了一种新颖的正交频分多址接入系统资源分配算法.通过最小化评估函数和取整操作,得到满足用户要求的数据速率和误码率,同时使系统总传输功率最小化.与传统遗传算法相比,该算法避免了注水过程,其计算复杂度大幅降低,且不受数据速率增长的影响.仿真实验表明该算法的收敛性和稳定性好;在相同的用户数据速率和误码率要求下,该算法得到的最小化传输功率随用户数增加较为缓慢,接近于解析法得到的理想性能;在平均比特信噪比相同的情况下,该算法能使实际系统误码率接近理论低值.

Combined the conventional genetic algorithm with the analytic method,a novel algorithm was proposed to allocate the source of orthogonal frequency division multiplex access system.Through minimizing the evaluation function and operating a rounding,the total transmitted power can be minimized while satisfying the bit rate requirement and bit error rate（BER） constraint of each user.Compared with the conventional genetic algorithm the proposed algorithm,can avoid the water-filling processing,so its computational complexity decreases sharply and is independent of the bit rate.The simulations indicate that the proposed algorithm has better convergence and stability than the conventional genetic algorithm.Under the same required user＇s bit rate and BER,the minimized transmitted power obtained by the proposed algorithm increases slowly with the number of the user,which is close to the performance of the analytic method.Given the same average bit signal to noise ratio,the proposed algorithm can successfully achieve the BER that is almost close to the theoretical minimum.