|
|
联合深度学习与FDSS的抑制PAPR研究
|
Abstract:
OFDM是5G物理层关键技术之一，其缺点是PAPR过高，容易导致功放效率下降并造成信号失真。如何抑制OFDM信号的PAPR对低功耗的物联网终端来说是一个重要问题。本文提出了一种联合深度学习与FDSS的抑制PAPR算法。仿真结果表明，所提算法对于多种调制方式及子载波个数配置均有很好的PAPR抑制效果。在峰值功率受限的条件下，采用所提算法能使信道的传输增益提升6 dB左右。
OFDM, one of the key techniques of the 5G physical layer, has the disadvantage of excessively high PAPR. The excessively high PAPR will lead to a decrease in power amplifier efficiency and cause signal distortion. How to suppress the PAPR of OFDM signals is an important problem for low-power Internet of Things terminals. This paper proposes a joint method combining deep learning and FDSS for PAPR suppression based on the PAPR suppression scheme of FDSS, and conducts simulation verification. The results show that the proposed joint method achieves excellent PAPR suppression performance in different modulation scenarios and different subcarrier numbers. Under the condition of peak power constraint, the proposed joint method can improve the transmission gain of the channel by about 6 dB.
| [1] | Nasarre, I.P., Levanen, T., Pajukoski, K., Lehti, A., Tiirola, E. and Valkama, M. (2021) Enhanced Uplink Coverage for 5G NR: Frequency-Domain Spectral Shaping with Spectral Extension. IEEE Open Journal of the Communications Society, 2, 1188-1204. https://doi.org/10.1109/ojcoms.2021.3082688 |
| [2] | 段向阳, 辛雨, 暴桐, 等. 一种高频场景候选波形方案[J]. 电子与信息学报, 2021, 43(1): 60-67. |
| [3] | 黄梦佳. OFDM系统中基于预畸变技术的峰均比抑制方案的研究[D]: [硕士学位论文]. 北京: 北京邮电大学, 2019. |
| [4] | 韩东升, 杨维, 刘薇. 一种改进的限幅滤波降低PAPR算法[J]. 北京邮电大学学报, 2014, 37(4): 44-48. |
| [5] | Chen, L., Pan, J., Zhang, J., Cheng, J., Xu, L. and Ye, N. (2023) FDSS-Based DFT-S-OFDM for 6G Wireless Sensing. Sensors, 23, Article No. 1495. https://doi.org/10.3390/s23031495 |
| [6] | Sahin, A., Hosseini, N., Jamal, H., Hoque, S.S.M. and Matolak, D.W. (2021) DFT-Spread-OFDM-Based Chirp Transmission. IEEE Communications Letters, 25, 902-906. https://doi.org/10.1109/lcomm.2020.3039222 |
| [7] | da Silva, B.S.d.C., Souto, V.D.P., Souza, R.D. and Mendes, L.L. (2024) A Survey of PAPR Techniques Based on Machine Learning. Sensors, 24, Article No. 1918. https://doi.org/10.3390/s24061918 |
| [8] | Wang, X., Jin, N. and Wei, J. (2021) A Model-Driven DL Algorithm for PAPR Reduction in OFDM System. IEEE Communications Letters, 25, 2270-2274. https://doi.org/10.1109/lcomm.2021.3076605 |
| [9] | Omar, M.S., Qi, J. and Ma, X. (2023) Mitigating Clipping Distortion in Multicarrier Transmissions Using Tensor-Train Deep Neural Networks. IEEE Transactions on Wireless Communications, 22, 2127-2138. https://doi.org/10.1109/twc.2022.3209188 |
| [10] | Chakravarty, S. and Kumar, A. (2023) PAPR Reduction of GFDM Signals Using Encoder-Decoder Neural Network (Autoencoder). National Academy Science Letters, 46, 213-217. https://doi.org/10.1007/s40009-023-01230-1 |
