Green DetNet: Computation and Memory efficient DetNet using Smart Compression and Training

TL;DR

This paper presents Green DetNet, a memory and computation-efficient MIMO detection neural network using incremental training and structured regularization, achieving significant reductions in memory and FLOPs while maintaining performance.

Contribution

Introduces an incremental training framework with structured regularization to optimize DetNet for reduced memory and computational costs.

Findings

98.9% reduction in memory requirement

81.63% reduction in FLOPs

Maintains BER performance despite compression

Abstract

This paper introduces an incremental training framework for compressing popular Deep Neural Network (DNN) based unfolded multiple-input-multiple-output (MIMO) detection algorithms like DetNet. The idea of incremental training is explored to select the optimal depth while training. To reduce the computation requirements or the number of FLoating point OPerations (FLOPs) and enforce sparsity in weights, the concept of structured regularization is explored using group LASSO and sparse group LASSO. Our methods lead to an astounding $98.9\%$ reduction in memory requirement and $81.63\%$ reduction in FLOPs when compared with DetNet without compromising on BER performance.