#ThinkSix - Validating a Machine-Learning Based Neural Receiver with 5G NR Multiple MIMO Signals

Using Machine Learning (ML) powered by Artificial Intelligence for signal processing tasks in wireless communication really is beginning to move from theory to practice. This video demonstrates the first hardware setup capable of validating the performance of a self-training neural receiver. Fully compliant 5G signals are transmitted and received with R&S equipment, to simulate two users and 2x2 MIMO communication. The received and synchronized signals are input to a prototype neural receiver from NVIDIA. The neural receiver replaces the conventional channel estimating, equalizing and de-mapping procedures in a receiver in a single processing step. Block error rate curves for the neural receiver show lower error rates compared to the data processing complexity for a self-training neural receiver, compared to conventional receiver algorithms.