联合星座扩展的预留子载波OFDM信号峰均比抑制算法

Tone reservation technique combined with active constellation extension for PAPR reduction of OFDM signal

针对现有预留子载波TR技术对无线OFDM信号峰均比(PAPR)抑制性能效率低,且难以同时兼顾峰均比抑制、误码率(BER)性能损失及带外频谱分量扩展的问题,提出一种联合智能梯度映射主动星座扩展ACE的预留子载波峰均比抑制ACE-TR算法,能以较低的复杂度同时对信号峰均比和接收端误码率性能进行联合优化,并在迭代过程中消除因限幅处理所导致的信号带外频谱分量再生;特别是,由于在优化迭代过程中可以对迭代参数进行自适应调整,能够有效提高算法的适用灵活性。对算法进行了全面深入的理论分析,推导了其可获得的PAPR抑制增益理论界和接收信号误码率性能理论值。理论分析与仿真表明,ACE-TR算法能以更快的收敛速度产生所需的削峰信号,并同时获得优异的峰均比抑制、误码率及带外功率谱性能。

To overcome the drawbacks of existing tone reservation（TR） techniques, an efficient TR algorithm combined with active constellation extension（ACE-TR） was proposed for reducing the peak-to-average power ratio（PAPR） of orthogonal frequency division multiplexing（OFDM） signal. By means of a specially designed iterative procedure and a joint optimization approach for PAPR and bit-error-rate（BER） performances, ACE-TR was able to obtain both an improved BER and minimized out-of-band interference while reducing the PAPR effectively. A comprehensive theoretical analysis was presented, and some important results including the bounds of achievable PAPR gain, BER bound, and maximum iteration number were derived. Specifically, by adaptively adjusting the iterative parameters in the optimization approach, more trade-off flexibility between PAPR reduction and BER performance can be offered to satisfy various design requirements. Simulations demonstrate that ACE-TR can dramatically decrease the number of required iterations to reach the desired PAPR with low computation complexity. In addition, the transmitted OFDM symbols using the proposed ACE-TR have less in-band distortion and lower out-of-band spectral regrowth than traditional TR algorithms.