Download PDFOpen PDF in browser

Sparsely Connected Neural Network for Massive MIMO Detection

EasyChair Preprint 376

6 pagesDate: July 24, 2018

Abstract

Deeping learning can achieve high parallelism and robustness, which is especially suitable for massive multiple-input multiple-output (MIMO) detection. There are already some well-developed deep learning models applied to MIMO detection, in which detection network is a typical representative model with excellent performance, but its complexity is high. This paper aims to simplify the detection network model, and the simplification runs through the entire data processing. This simplification includes three improvements. First, the number of inputs is reduced to simplify inputs; Second, the network connection structure is simplified by changing network from full connectivity to sparsely connectivity and reducing the number of network layers by half. Third, the loss function optimizes to avoid irreversible problems with the matrix. Base on the above improvements, the complexity of the network is reduced from O(64n2) to O(3n). The simulation results indicates that the proposed structure has better performance than the existing detection network.

Keyphrases: MIMO detection, Sparsely Connected Neural Network, deep learning

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:376,
  author    = {Guili Gao and Chao Dong and Kai Niu},
  title     = {Sparsely Connected Neural Network for Massive MIMO Detection},
  doi       = {10.29007/rjdc},
  howpublished = {EasyChair Preprint 376},
  year      = {EasyChair, 2018}}
Download PDFOpen PDF in browser