Adaptive Neural Signal Detection for Massive MIMO

Understand

Symbol detection for Massive Multiple-Input Multiple-Output (MIMO) is a challenging problem for which traditional algorithms are either impractical or suffer from performance limitations.

• Several recently proposed learning-based approaches achieve promising results on simple channel models (e.g., i.i.d.
• Gaussian).
• However, their performance degrades significantly on real-world channels with spatial correlation.