Noise, channel and message identification on MIMO channels with general noise

In this paper, we consider the decodification of messages in a noisy MIMO environment. Initially, two steps are performed: firstly the identification of the noise characteristics by using uncertainty quantification methods. Then, the channel matrix is identified by decoding known messages. In a last step, we show that the method is effective to decode generic messages and we analyze the error rate. We consider alphabets going from 2 to 128 symbols.