Sensitivity analysis of the channel estimation deviation to the MAP decoding algorithm

Abstract

As a necessary input parameter for maximum a-posteriori(MAP) decoding algorithm, SNR is normally obtained from the channel estimation unit. Corresponding research indicated that SNR estimation deviation degraded the performance of Turbo decoding significantly. In this paper, MAP decoding algorithm with SNR estimation deviation was investigated in detail, and the degradation mechanism of Turbo decoding was explained analytically. The theoretical analysis and computer simulation disclosed the specific reasons for the performance degradation when SNR estimation was less than the actual value, and for the higher sensitivity of SNR estimation to long-frame Turbo codes.