A Comparison of Decoding Latency for Block and Convolutional Codes⋆

In this paper we compare the decoding latency, i.e., the delay between the time a channel symbol is received and the time it is decoded, of block and convolutional codes. In particular, we compare low density parity check (LDPC) block codes with iterative message-passing decoding to convolutional codes with Viterbi decoding and stack sequential decoding. On the basis of simulations, we show that, for a code rate of 1/2, a target bit error rate of 10, and an allowed latency of up to approximately 2000 information bits, convolutional codes with stack sequential decoding require a smaller signal-to-noise ratio (SNR) than LDPC codes with iterative message-passing decoding. For larger allowed latencies, the advantage switches to LDPC codes.