Joint Model and Data-Driven Receiver Design for Data-Dependent Superimposed Training Scheme With Imperfect Hardware

Data-dependent superimposed training (DDST) scheme has shown the potential to achieve high bandwidth efficiency, while encounters symbol misidentification caused by hardware imperfection. To tackle these challenges, a joint model and data driven receiver is proposed in this paper. Specifically, based on conventional linear model, least squares (LS) estimation zero forcing (ZF) equalization are first employed extract initial features for channel detection. Then, shallow neural networks, named CE-Net SD-Net, developed refine detection, where imperfect modeled as nonlinear function utilized train networks approximate it. Simulation results show that compared with minimum mean square error (MMSE) scheme, one effectively suppresses achieves similar or better bit rate (BER) performance without second-order statistics about noise.