Early-termination of decoding convolutional codes

A decoding method, comprising:

• receiving, at an input of a decoder, a sequence of softbits presumed to correspond to a convolutionally-encoded codeword;
• executing, by the decoder for the received sequence of softbits, a Viterbi-based algorithm for:
• determining a Maximum Likelihood (ML) survivor path in a trellis representation of the codeword;
• determining whether the presumed convolutionally-encoded codeword meets an early-termination criteria;
• aborting execution of the Viterbi-based algorithm for the received sequence of softbits to stop the decoding process if the presumed convolutionally-encoded codeword meets the early-termination criteria irrespective of a number of survivor path update iterations that have already been performed such that the received sequence of softbits are not further processed as part of a decoder operation and the ML survivor path is not decoded;
• and
• continuing the decoding method if the presumed convolutionally-encoded codeword fails to meet the early-termination criteria,
• wherein the early-termination criteria indicates that the codeword is ill-formed based upon an assessed likelihood of the received sequence of softbits being a correctly encoded codeword,
• wherein the early-termination criteria includes a heuristic classification based on a distribution of survivor-path initial states, and
• wherein the act of determining whether the presumed convolutionally-encoded codeword meets the early-termination criteria includes determining a metric based upon the distribution of survivor-path initial states, and determining whether the codeword is ill-formed based upon comparison of the metric to a threshold, thereby classifying between correctly-formed and ill-formed codewords.