Neural network error correcting decoders for block and convolutional codes
William R. Caid, Robert W. Means · 2002
The use of neural networks as error correcting decoders is described. It is shown that the neural networks may offer advantages in electronic countermeasure (ECM) environments in which the convolutional design assumptions of additive white Gaussian noise (AWGN) and a binary symmetric channel (BSC) are violated. Some results of preliminary studies and benefits of the neural-based decoder approach are discussed.>