Turbo decoding over a two-state Markov channel

A decoding system comprising of a turbo decoder and a burst detector is proposed for a two-state, additive white Gaussian noise, Markov channel in which one state represents a good state with high E_b/N₀ and the other state represents a bad state or burst state with low E_b/N₀. As the Gaussian noise model allows soft information from the channel to be used, it enables an improved performance over the commonly used binary-input/binary-output channel model. In the proposed decoding structure, the burst detector is employed to estimate the probability of each channel state and pass it on to the turbo decoder, so that this extra information assists the turbo decoder to perform a more effective decision for each received symbol by adjusting the likelihood function property in accordance with the channel states.