AI Deep Learning Simulating Wireless Communications

SIMULATION 1: Gradient Computation

a) Setting up end-to-end communication system

It is setting up a simple communication system that transmit bits modulated as QAM symbols over an AWGN channel. Transmit bits modulated as QAM symbols over an AWGN channel to make the constellation trainable.

b) Use of Stochastic Gradient Descent on the received QAM-AWGN

It illustrates the constellation after applying the gradient application: Optimization of the constellation through Stochastic Gradient Descent using the optimizer Adam. The gradients can be applied to the trainable weights to update them.

c) Computing and Plotting the Evaluated Trained Model QAM-AWGN-SGD

SIMULATION 2: MIMO Transmissions over a Flat-Fading Channel

a) System model with requirements

b) Transmitted Received Constellations

Based on the above design, the received constellations MIMO Flat-Fading is illustrated as:

c) BER simulations using a Keras model

Coming Soon More Simulations...