针对基于遗传算法(GA)的部分传输序列(PTS)方法在降低正交频分复用(OFDM)系统峰均比(PAPR)时存在避免早熟收敛和降低算法复杂度两项指标不能兼顾的问题,提出分段替换的降低OFDM峰均比算法。通过设置合理的门限值,减少不必要的搜索运算,降低算法复杂度;利用克隆种群和记忆种群相结合的分段替换染色体策略,提高优质种群利用率,加快收敛速度的同时避免早熟收敛。仿真结果表明,合理的门限值和分段替换染色体策略可以优化降低峰均比算法的性能。
Based on segment replacement scheme, an improved genetic algorithm (GA) partial transmits sequence (PTS) is proposed. It can achieve a trade-off between overcoming premature and complexity reduction in the process of reducing peak-average-power ratio (PAPR) in an orthogonal frequency division multiplexing (OFDM) system. Appropriate threshold can reduce the unnecessary search operation and reduce the complexity of the algorithm. Combining the clone population and the memory population, the segment replacement strategy is proposed not only to improve the utilization of fine species, but also to avoid premature convergence. Simulation results illuminate that the reasonable threshold and segment replacement strategy can improve the optimize performance of the algorithm.