ANNIXIBLE

The goal of this project is to use machine learning at the receiver side of the communication system by creating a neural network decoder for a binary error-correcting code. The digital system used in this project will consists of five stages: a source that generates bits in packets of k, an encoder that reads those k bits and producing n bits, a binary symmetric channel, an error-correcting decoder containing the neural network, and a receiver of the k decoded bits. Such a decoder reduces the complexity and improves the flexibility of the digital receiver.