摘要
广义频分复用(Generalized Frequency Division Multiplex,GFDM)即将成为未来5G多载波调制系统的备选方案,针对其峰均功率比(Peak-to-Average Power Ratio,PAPR)较高的缺点,提出了一种基于DHT预编码矩阵的Cos分布非线性压扩抑制PAPR的联合算法。利用DHT预编码技术,通过预编码矩阵乘法操作对频域信号进行处理,降低同一时隙不同载波间符号的非周期自相关性,避免了子载波调制数据时相位的一致性,能够消除GFDM调制阶段的自干扰与信道干扰,降低GFDM信号系统的瞬时功率,PAPR得到抑制。为了达到更优的抑制效果且对系统误码率(Bit Error Rate,BER)影响较小,对预编码后的序列进行Cos非线性压扩以降低各信号的幅值。理论分析和仿真结果表明,提出的算法比仅采用DHT预编码或压扩法具有更低的PAPR,抑制效果更明显。
Generalized Frequency Division Multiplexing(GFDM)is about to become an alternative scheme for future 5G multicarrier modulation systems,but its Peak-to-Average Power Ratio(PAPR)is still high,therefore,a joint algorithm for PAPR reduction based on DHT precoding matrix and Cos distribution nonlinear companding is proposed.Firstly,DHT precoding technology is used to process frequency domain signals through precoding matrix multiplication operation to reduce the aperiodic autocorrelation of symbols between different carriers in the same time slot,avoid the phase consistency of subcarrier modulation data and it can eliminate the self-interference and channel interference in the GFDM modulation phase,thus reducing the instantaneous power of GFDM signal system and suppressing PAPR.In order to attain better suppression result and lower influence on the system Bit Error Rate(BER),Cos nonlinear companding is performed after the pre-coding transform to reduce the amplitude of each signal.Theoretical analysis and simulation results show that the proposed algorithm has lower PAPR and more obvious suppression effect than DHT precoding or companding method.
作者
谷玉婷
刘春刚
田欣欣
谭畅
王钢
杨珊珊
GU Yuting;LIU Chungang;TIAN Xinxin;TAN Chang;WANG Gang;YANG Shanshan(College of Middle Combustion Engineering,Hebei Normal University,Shijiazhuang 050024,China;School of Electronic and Information Engineering,Harbin Institute of Technology,Harbin 150001,China)
出处
《无线电工程》
北大核心
2022年第10期1726-1733,共8页
Radio Engineering
基金
国家自然科学基金(62071167)。
关键词
广义频分复用
峰均功率比
DHT预编码
Cos分布非线性压扩
generalized frequency division multiplexing
peak-to-average power ratio
DHT precoding
Cos distribution nonlinear companding