Joint Compensation of CFO and IQ Imbalance in OFDM Receiver: A Deep Learning Based Approach

Abstract

Due to the technical and cost limitations, wireless systems suffer from various hardware impairments, including phase noise, power amplifier nonlinearity, carrier frequency offset and in-phase and quadrature-phase imbalance. These impairments can highly degrade the physical layer performance and are usually compensated separately by using model-based signal processing techniques. However, due to the high carrier frequency and large bandwidth of 5G new radio, the coupling effects between different impairments are highly aggravated, which greatly degrades the performance of individual compensation modules for different impairments. In this paper, we propose a deep learning-based method, which jointly addresses the hardware impairments directly from the received data. Specifically, we focus on carrier frequency offset and in-phase and quadrature-phase imbalance, and propose a deep neural network with multiple parallel subnets for joint compensation. Numerical results show that the proposed method outperforms the conventional method using separate compensation modules in practical signal-to-noise ratio regions, and the performance improvement further increases when the cyclic prefix length or the pilot length is limited.