DeepReceiver: A Deep Learning-Based Intelligent Receiver for Wireless Communications in the Physical Layer

A canonical wireless communication system consists of a transmitter and receiver. The information bit stream is transmitted after coding, modulation, pulse shaping. Due to the effects radio frequency (RF) impairments, channel fading, noise interference, signal arriving at receiver will be distorted. needs recover original from distorted signal. In this article, we propose new model, namely DeepReceiver, that uses deep neural network replace traditional receiver's entire recovery process. We design one-dimensional convolution DenseNet (1D-Conv-DenseNet) structure, in which global pooling used improve adaptability different input lengths. Multiple binary classifiers are final classification layer achieve multi-bit recovery. also exploit DeepReceiver for unified blind reception multiple modulation coding schemes (MCSs) by including samples corresponding MCSs training set. Simulation results show proposed performs better than step-by-step serial hard decision terms error rate under influence various factors such as noise, RF multipath cochannel dynamic environment, MCSs.