An in-service SNR estimation scheme associated with hybrid ARQ over correlated frequency-flat Rayleigh fading channel

We focus on the SNR estimation scheme associated with hybrid ARQ over correlated frequency-flat Rayleigh fading channel. Due to the existing correlation among the fading amplitude within neighboring code blocks in the hybrid ARQ, accumulative observed noisy signal sequence (AONSS) or the differential observed noisy signal sequence (DONSS) could be utilized in the SNR estimation associated with hybrid ARQ scheme. The corresponding Cramer-Rao lower bound is derived. It is indicated that the ML SNR estimation can benefit from the AONSS in the correlated fading channel, especially for the slow fading cases. By using the alternative DONSS sequences, an in-service maximum likelihood SNR estimation technique is also proposed. Finally, both the derived ML SNR estimation and the applicability of the proposed in-service SNR estimation techniques are validated by means of numerical analysis and through simulation results.