Abstract:Deep-unfolding neural networks (NNs) have received great attention since they achieve satisfactory performance with relatively low complexity. Typically, these deep-unfolding NNs are restricted to a fixed-depth for all inputs. However, the optimal number of layers required for convergence changes with different inputs. In this paper, we first develop a framework of deep deterministic policy gradient (DDPG)-driven deep-unfolding with adaptive depth for different inputs, where the trainable parameters of deep-unfolding NN are learned by DDPG, rather than updated by the stochastic gradient descent algorithm directly. Specifically, the optimization variables, trainable parameters, and architecture of deep-unfolding NN are designed as the state, action, and state transition of DDPG, respectively. Then, this framework is employed to deal with the channel estimation problem in massive multiple-input multiple-output systems. Specifically, first of all we formulate the channel estimation problem with an off-grid basis and develop a sparse Bayesian learning (SBL)-based algorithm to solve it. Secondly, the SBL-based algorithm is unfolded into a layer-wise structure with a set of introduced trainable parameters. Thirdly, the proposed DDPG-driven deep-unfolding framework is employed to solve this channel estimation problem based on the unfolded structure of the SBL-based algorithm. To realize adaptive depth, we design the halting score to indicate when to stop, which is a function of the channel reconstruction error. Furthermore, the proposed framework is extended to realize the adaptive depth of the general deep neural networks (DNNs). Simulation results show that the proposed algorithm outperforms the conventional optimization algorithms and DNNs with fixed depth with much reduced number of layers.

An Iterative BP-CNN Architecture for Channel Decoding.

Exploiting Noise Correlation for Channel Decoding with Convolutional Neural Networks

An Iterative BP-CNN Decoder for Optical Fiber Communication Systems.

A denoiser for correlated noise channel decoding: Gated-neural network

A Neural Network-Based Compressive LDPC Decoder Design over Correlated Noise Channel

A UnetNND-BP Architecture for Channel Decoding Under Correlated Noise

Co-Channel Interference And Colored Noise Mitigation: An Iterative Structure With Convolutional Neural Network

Channel Noise Optimization of Polar Codes Decoding Based on a Convolutional Neural Network

A Model-Driven Quasi-ResNet Belief Propagation Neural Network Decoder for LDPC Codes.

A Novel Decoding Scheme for Polar Code Using Convolutional Neural Network

A Novel Neural Network Denoiser for BCH Codes.

Performance Evaluation of Channel Decoding with Deep Neural Networks.

Convolutional Neural Network-aided Bit-flipping for Belief Propagation Decoding of Polar Codes

Dark and Bright Channel Prior Embedded Network for Dynamic Scene Deblurring

Research on Blind Recognition Algorithm of Channel Coding Based on One-Dimensional Convolutional Neural Network Under the Low SNR Regime

DDPG-Driven Deep-Unfolding with Adaptive Depth for Channel Estimation with Sparse Bayesian Learning

Doubly Residual Neural Decoder: Towards Low-Complexity High-Performance Channel Decoding

Multinoise-type Blind Denoising Using a Single Uniform Deep Convolutional Neural Network.

Iteration-Aware Dynamic Quantization Neural Belief Propagation Decoder

A Novel Convolutional Neural Network Architecture of Deep Joint Source-Channel Coding for Wireless Image Transmission

Learning Channel-Wise Interactions for Binary Convolutional Neural Networks.