基于神经网络的梳状谱信号峰均比优化技术  

PAPR Reduction of Comb Spectrum Interference Signals

作　　者：安坤 杨腾飞 AN Kun;YANG Teng-fei(Xidian University,Xi'an 710071,China)

机构地区：[1]西安电子科大学,陕西西安710071

出　　处：《中国电子科学研究院学报》2020年第6期566-572,共7页Journal of China Academy of Electronics and Information Technology

摘　　要：本文针对梳状谱干扰信号存在峰均比较大的问题,梳理分析了梳状谱信号峰均比与限幅率和干扰信号数量之间的关系。同时,提出一种基于神经网络学习的峰均比改善方法。该方法利用神经网络能够自适应的选择合适的限幅率对梳状谱干扰信号进行限幅操作,在有效改善峰均比的同时能够保持信号原有的谱特性。最后,通过计算机仿真结果验证,本文算法能有效改善梳状谱干扰的峰均比。For the interference signal of comb spectrum,this article has the problem of large PAPR The relationship between the PAPR of the comb spectrum interference signals and the clipping rate and the number of interfering signals was analyzed.At the same time,a reduced method of PAPR based on neural network learning is proposed.This method can adaptively select an appropriate limiting rate to perform limiting operations on the comb spectrum interference signal through a neural network,which effectively reduces the PAPR while maintaining the original spectral characteristics of the signal Finally,computer simulation results verify that the algorithm in this paper can effectively reduce PAPR the of comb spectrum signals.

关 键 词：梳状谱干扰信号 峰均比改善 限幅法 神经网络学习 

分 类 号：TN97[电子电信—信号与信息处理]