基于SLM-PTS算法融合的NC-OFDM峰均比优化  

PAPR optimization based on SLM and PTS algorithms in NC-OFDM systems

作　　者：周杰[1,2] Esono Mikue Bernardo Esono 王学英 周惠婷 罗宏 ZHOU Jie;Esono Mikue Bernardo Esono;WANG Xueying;ZHOU Huiting;LUO Hong(School of Electronic and Information,Nanjing University of Information Science and Technology,Nanjing 210044,China;Library,Nanjing University of Information Science and Technology,Nanjing 210044,China)

机构地区：[1]南京信息工程大学电子与信息工程学院,江苏南京210044 [2]南京信息工程大学图书馆,江苏南京210044

出　　处：《电信科学》2022年第7期63-74,共12页Telecommunications Science

基　　金：国家自然科学基金资助项目(No.61971167,No.62101275,No.62101274);江苏省信息与通信工程优势学科建设项目。

摘　　要：基于非连续正交频分复用(non-continuous orthogonal frequency division multiplexing,NC-OFDM)模型,提出和研究了选择映射(selected mapping,SLM)算法和部分传输序列(partial transmit sequence,PTS)算法,及其SLM-PTS融合优化技术,设计了融合模型和改进流程。仿真结果与其他文献方法进行了对比,验证了SLM-PTS的融合具有优秀的峰值平均功率比(peak to average power ratio,PAPR)降低能力,但缺点是算法实现复杂度过高。因此,又进一步提出了互补型映射和限幅的联合算法(SLM-Clipping)融合解决方案,并利用深度学习方法建立PAPRnet模型。仿真结果验证了此算法对NC-OFDM系统具有PAPR良好的抑制效果,而且能够提高仿真运算效率。Based on the non-continuous orthogonal frequency division multiplexing(NC-OFDM)model,a fusion optimization technology based on selected mapping(SLM)algorithm and partial transmit sequence(PTS)algorithm was proposed,and a system model of fusion technology was designed.Through simulation comparison with other literature methods,it was verified that the SLM-PTS fusion technology had excellent peak to average power ratio(PAPR)reduction ability,but the algorithm implementation complexity was too high.Therefore,a complementary SLM-Clipping fusion solution was proposed,and the deep learning method PAPRnet model was construted.The simulation results verif that prove the effectiveness of the method,the algorithm has an excellent PAPR suppressed effect on the NC-OFDM system,and greatly improves the computational efficiency.

关 键 词：NC-OFDM 选择映射算法 部分传输序列算法 融合算法 深度学习算法 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]