Research Article
Complex-Valued Pipelined Recurrent Neural Network for Transmitter Distortions Compensation in High-Throughput Satellite Communication
@INPROCEEDINGS{10.1007/978-3-030-69072-4_36, author={Changzhi Xu and Yi Jin and Li Yang and Li Li and Mingyu Li and Zhenxin Cao}, title={Complex-Valued Pipelined Recurrent Neural Network for Transmitter Distortions Compensation in High-Throughput Satellite Communication}, proceedings={Wireless and Satellite Systems. 11th EAI International Conference, WiSATS 2020, Nanjing, China, September 17-18, 2020, Proceedings, Part II}, proceedings_a={WISATS PART 2}, year={2021}, month={2}, keywords={Digital predistortion (DPD) Complex-valued pipelined recurrent neural network (CPRNN) I/Q imbalance Power amplifiers (PAs) Satellite communication}, doi={10.1007/978-3-030-69072-4_36} }- Changzhi Xu
Yi Jin
Li Yang
Li Li
Mingyu Li
Zhenxin Cao
Year: 2021
Complex-Valued Pipelined Recurrent Neural Network for Transmitter Distortions Compensation in High-Throughput Satellite Communication
WISATS PART 2
Springer
DOI: 10.1007/978-3-030-69072-4_36
Abstract
With the continuous development of satellite communication system towards the direction of high frequency band, large capacity and high spectral efficiency transmission, the signals processed by these new technologies have many characteristics, such as ultra-high bandwidth and higher peak-to-average power ratio (PAPR), etc., which puts forward a great challenge for the spaceborne transmitter used in the satellite communication system. In view of the above requiremes, a novel digital predistortion (DPD) model based on complex-valued pipelined recurrent neural network (CPRNN) for joint compensation of wideband spaceborne transmitter is proposed in this paper. Once the CPRNN model is constructed, the complex-valued real time recurrent learning (CRTRL) algorithm is used for the CPRNN model training. Here, the CRTRL algorithm is derivated in detail based on the real-valued RTRL algorithm. The imperfect transmitters based on a GaN PA excited by the 400-MHz 64-amplitude/phase-shift keying (64APSK) signals was employed to verify the compensation performance of the proposed models. The simulation and experimental results show that the proposed CPRNN DPD model can achieve better linearization performance for the nonlinear transmitter with imperfect RF impairments.
