Performance of Raptor Codes on the BIAWGN Channel in the Presence of SNR Mismatch

Abstract

Accurate estimation of the channel signal to noise ratio (SNR) is essential for belief propagation (BP) decoding to operate optimally. Incorrect estimation of the channel SNR is known as SNR mismatch and can lead to serious degradation in BP decoding performance especially when a code is operating near its decoding threshold. We analyze the asymptotic performance of Raptor codes under SNR mismatch on the binary input additive white Gaussian noise (BIAWGN) channel using discretized density evolution (DDE). We provide the decoding thresholds of Raptor codes for a wide range of SNR mismatch values. Our results show that overestimation of channel SNR is slightly more detrimental than underestimation for lower levels of SNR mismatch, while, underestimation becomes more detrimental as the mismatch increases. Finally, we use DDE-based optimization to design SNR mismatch tolerant output degree distributions.