Companding is a well-known technique for the peak-to-average-power ratio (PAPR) reduction of orthogonal frequency division multiplexing (OFDM) signals. Piecewise linear companding (PLC) scheme, compared with other companding schemes, can effectively reduce PAPR with much lower complexity, while achieving improved bit-error-rate (BER) performance. However, when the PAPR target value for companded signals is relatively small, such as 4 dB for 16-QAM or 5 dB for 64-QAM, there is a clear error floor of PLC in high SNR region for high-order QAM modulation. Focusing on this problem, we first analyze the reason for the error floor of PLC. Based on the analysis, we propose a generalized PLC scheme to reduce the distortion in decompanded signals with the relaxation of the constraint on the conservation of the average signal power. Finally, we formulate the optimization problems to obtain the parameters of the companding function in the pursuit of a preferable tradeoff between the BER and power spectral density performance under the constraint on PAPR. Simulation results verify that the proposed generalized PLC can significantly improve the BER performance while maintaining the same PAPR performance as PLC.

中文翻译：

---

用于降低 OFDM 系统中 PAPR 的广义分段线性压扩变换

压扩是一种众所周知的技术，用于降低正交频分复用 (OFDM) 信号的峰均功率比 (PAPR)。分段线性压扩 (PLC) 方案与其他压扩方案相比，可以有效降低 PAPR，同时降低复杂度，同时实现更高的误码率 (BER) 性能。然而，当压扩信号的 PAPR 目标值相对较小时，例如 16-QAM 的 4 dB 或 64-QAM 的 5 dB，对于高阶 QAM 调制，PLC 在高 SNR 区域存在明显的误差基底。针对这个问题，我们首先分析PLC的error floor的原因。在分析的基础上，我们提出了一种广义的 PLC 方案，以减少解压缩信号中的失真，同时放宽对平均信号功率守恒的约束。最后，我们制定优化问题以获得压扩函数的参数，以在 PAPR 的约束下追求 BER 和功率谱密度性能之间的最佳折衷。仿真结果验证了所提出的广义 PLC 可以显着提高 BER 性能，同时保持与 PLC 相同的 PAPR 性能。