Turbo decoding for high spectral efficiency satellite communications

The problem of efficient utilization of the frequency spectrum for coded satellite systems is investigated; one which results as a consequence of highly crowding adjacent channels. An analytical characterization of the resulting interference channel is introduced based on concepts in statistical decision theory and is then exploited for interference cancellation. A joint implementation of MMSE interference cancellation and forward error control (FEC) decoding is considered where soft-input soft-output (SISO) modules are used along with the iterative structure. It is shown that, for example, one can operate a satellite system with convolutionally encoded QPSK modulation that uses practical pulse shaping at the channel spacing valve of 0.75 of the symbol rate, or bandwidth efficiency level of 2.67 bits-per-second/Hz, with minimum additional energy requirement. (This corresponds to a spectral efficiency improvement of 55% compared with a conservative baseline system.)