A model-driven deep learning based multi-user joint decoding method for a cell-free massive MIMO system

By designing a low-complexity analog beam combination matrix and a MIMO-LDPC-Net decoding network in a large-scale MIMO system, the decoding performance problem in multi-user MIMO interference channels is solved, and the signal-to-noise ratio and decoding error probability are significantly improved.

CN121125017BActive Publication Date: 2026-07-24SOUTHEAST UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SOUTHEAST UNIV
Filing Date
2025-09-08
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

In large-scale machine-type communication scenarios, existing technologies are unable to effectively solve the channel interference problem in multi-user MIMO interference channels. Traditional MMSE receivers have high computational complexity, and the spatial degrees of freedom of the analog combination matrix are not fully utilized, affecting decoding performance.

Method used

A low-complexity analog beam combination matrix optimization method is designed using a model-driven deep learning approach. Multi-user joint decoding is achieved through the MIMO-LDPC-Net decoding network. Heuristic algorithms are combined to optimize the analog combination matrix, and the MIMO-LDPC-Net decoding network is constructed to improve decoding performance using deep learning.

Benefits of technology

In partially connected MIMO channels, the MIMO-LDPC-Net algorithm exhibits superior performance gains compared to the multi-user joint LDPC decoding algorithm across the entire signal-to-noise ratio range. The heuristic algorithm-assisted joint decoding algorithm provides a 1dB signal-to-noise ratio gain under high signal-to-noise ratio conditions, significantly improving the decoding error probability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121125017B_ABST
    Figure CN121125017B_ABST
Patent Text Reader

Abstract

This invention discloses a multi-user joint decoding method for cellular-free massive MIMO systems based on model-driven deep learning. In the simulation part of data decoding, an optimization problem of the simulation combination matrix is ​​established with the objective of maximizing the achievable rate of the cellular-free massive MIMO system. Based on a heuristic algorithm, the optimization objective is simplified, and the optimization problem is decomposed into multiple sub-problems. The single AP simulation precoding matrix optimization problem is solved one by one to obtain the optimal simulation combination matrix. The multi-user joint LDPC decoding algorithm is extended to the cellular-free massive MIMO scenario. Multi-user joint LDPC decoding is implemented based on the MIMO-LDPC-Net decoding network. In this network, the input layer generates initial values ​​as input to the hidden layer, which is composed of stacked modules consisting of a first sub-layer, a second sub-layer, and a MIMO detection layer. The output layer generates the log-likelihood ratio of the variable nodes. Experiments show that the heuristic algorithm + AI-assisted joint decoding algorithm has a 1dB signal-to-noise ratio gain compared to the traditional algorithm.
Need to check novelty before this filing date? Find Prior Art