.png){kind=icon}
Return
Model-driven learning-based message passing algorithm for MIMO systems
.png){kind=icon}
PRE
AI
DOI：10.1007/s11235-026-01451-x.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Deep learning (DL) has shown potential in solving the detection problem for multiple-input multiple-output (MIMO) systems with a better performance-complexity trade-off. In this paper, we propose a model-driven DL-based approximate message passing (AMP) detection algorithm for massive MIMO systems. We propose a method that adopts learnable parameters, which can be effectively used to enhance the performance of the existing AMP algorithm. The proposed method demonstrates faster convergence, requiring fewer iterations to achieve the target bit error rate (BER) performance. The proposed method also shows its robustness in generalization, as it is trained at a single signal-to-noise ratio (SNR) while being successfully tested over a range of SNR values. Furthermore, we extend our proposed method to extract soft information for a coded MIMO system with joint iterative estimation. Simulation results investigated in this paper reveal that the proposed method has notable performance improvement with much lower complexity compared to existing methods.
Keywords:
Approximate message passing
Deep learning
Massive MIMO
Signal detection
Journal
IF:
2.3
Papers:
161
Citations:
2.0K
Organization
Cited Papers
.png){kind=icon}
ShareSaveShareSaveShareSave
ShareSave
View more