A Parallel Decoding Scheme for the Multiple Stack Algorithm

Abstract

The multiple stack algorithm (MSA) is an efficient algorithm for erasurefree sequential decoding of long constraint length convolutional codes. The basic idea of the MSA is to make use of higher order stacks for blocks that require an excessive number of computations to be decoded. Unfortunately. Those blocks are frequently decoded with poor error performance if the decoding process is terminated with the computational linit, C_lim. We propose in this paper a decoding scheme to impeove the error performance of the MSA  with the use of parallel decoding. It is found by extensive computer simulations that the bit error rate of the  MSA can be further improved with our proposed scheme. Parallel decoding scheme for the MSA appcars to be an interesting scheme for decoding convolutional codes when low error probabilities are required.