AI for Wireless Communications with MATLAB

Day 1 of 1

Data Generation, Importing and Management

Objective: Use Wireless Waveform Generator app to interactively generate 5G NR standard-compliant waveforms. Configure and run 5G channel models to generate dataset. Use parallel for-loops to speed up computations. Use signal datastores to import and organize signal data in MATLAB.

• Wireless Waveform Generator App
• Generate and visualize 5G NR standard-compliant waveforms
• Configure cluster delay line MIMO fading channel
• Generate dataset for training and testing deep networks
• Run for-loops in parallel
• Import data with signal datastores

Creating and Training Networks

Objective: Create and train an autoencoder deep learning network to compress channel estimate for a CSI feedback system. Use Experiment Manager app to perform parameter sweep and manage results.

• Create training, validation and testing sets
• Create and visualize deep learning networks
• Use GPU to train deep learning networks
• Experiment Manager app

Running Python from MATLAB

Objective: Run python commands and scripts from within MATLAB and return the output variables to the MATLAB workspace. Create and train a PyTorch autoencoder-based neural network for massive MIMO CSI feedback.

• Set up the Python environment in MATLAB
• Execute Python commands from MATLAB
• Execute Python scripts from MATLAB
• Create and train deep learning networks in Python