A Bound on Decoding Error Probability in a Noisy Multiple-Access Channel

We derive upper bounds on the decoding error probability in a multiple-user channel with concentrated and impulse noise and white Gaussian noise. The concentrated noise is defined by a sum of harmonics with random frequencies and phases. The impulse noise is modeled by a Poisson stream of short pulses. All the users are independent, using common-format compound signals with frequency and phase jumps. For each receiver, the signals of other stations are treated as interference. The decision procedure includes correlative reception with subsequent decoding using a continuous output.