Implementation of sampling-based turbo decoder for noncoherent BFSK under Gaussian noise

This paper is concerned with the mathematical analysis for the sampling-based metric calculation for turbo decoding of noncoherent BFSK (NCBFSK) in additive white Gaussian noise (AWGN) channels, towards application in tactical applications. It is known that for NCBFSK, two decision statistics having Rician and Rayleigh characteristics must be accounted for in the decoding process. Therefore, this paper clarifies the means of using the two NCBFSK demodulator outputs so that sampling-based turbo decoding can be realized in simulation. Results for the simulations of the named waveform performances are shown for turbo coding rates 1/3, 2/3 and 3/4 under AWGN and a comparative assessment is presented.