Soft output decoding apparatus and method for convolutional code

A soft output decoding apparatus for convolutional codes, comprising:

• probability computing means for finding probability information in each transition state of the convolutional codes;
• probability storage means for storing the probability information as found by the probability computing means in a recording medium;
• and
• soft output computing means for finding a soft output using the probability information stored in said recording medium;
• wherein:
• the probability storage means stores the probability information in an amount not less than a truncated length on said recording medium;
• an updating of the probability information within the truncated length by the probability storage means and the computation of the soft output outside the truncated length by the soft output computing means are carried out in parallel;
• wherein said probability computing means comprises:
• first computing means for computing a first probability as determined by a code output pattern and a received value;
• second computing means for chronologically computing a second probability from the start of encoding up to each state for every received value, based on the first probability;
• and
• third computing means for computing a third probability for every received value from the truncation state up to each state in a sequence reversed from the chronological sequence, based on the first probability;
• and wherein the probability information updated within the truncated length of the probability storage means is the third probability information.

A soft output decoding method for convolutional codes comprising:

• a first step of finding probability information at each transition state of the convolutional codes;
• a second step of storing the probability information in a recording medium in a length not less than a truncated length;
• and
• a third step of finding a soft output using probability information stored in the recording medium in second step;
• wherein:
• an updating of the probability information within the truncated length in the second step is performed in parallel with the computation of the soft output outside the truncated length in the third step.

A soft output decoding method for convolutional codes, comprising:

• a first probability computing step of sequentially computing, for each received value, a first probability as determined by an output pattern of a code and a received value;
• a second probability computing step of computing, for each received value, a second probability of reaching each state in a direction along a time axis, based on the first probability;
• a third probability computing step of computing, for each received value, a third probability of reaching each state in a direction retrograding down the time axis, as from a preset reference time point, based on the first probability;
• and
• a soft output computing step of computing a soft output based on the first, second and third probabilities;
• wherein
• in the third probability computing step, the first probability is partitioned in terms of the pre-set truncated length as a unit, the reference time point is set along the time axis direction as from said truncated length, the first probability is processed simultaneously in parallel by plural channels, at least in terms of a length corresponding to a truncated length as from each reference time point as a unit, to compute the third probabilities by plural channels, the third probability corresponding to said truncated length is selected from the computed plural channels of the third probability, and the third probability corresponding to the received value is outputted;
• and wherein in the first probability computing step, the first probability is transiently held in probability storage means and sequentially read out and outputted in a sequence corresponding to each processing in the second and third probability computing steps and in the soft output computing step;
• and the third probability is transiently held in a probability storage means during the third probability computing step and sequentially read out and outputted in a sequence corresponding to the processing in the soft output computing step.

A soft output decoding apparatus for convolutional codes, comprising:

• an input terminal which inputs a priori probability values and a received value;
• a controller which generates control signals;
• a first computation storage circuit which computes and stores a first probability for each received value, in accordance with the a priori probability values and the received value from the input terminal and a control signal from the controller;
• a second computation storage circuit which computes and stores a second probability from an encoding state chronologically to each state from one received value to another, in accordance with the first probability from the first computation storage circuit and a control signal from the controller;
• a third computation storage circuit which computes a third probability for every received value from a pre-set truncated length up to each state in a reverse chronological sequence, in accordance with the first probability from the first probability computation storage circuit and a control signal from the control circuit;
• a soft output computation circuit which computes soft output values outside the truncated length in parallel with an updating of the probability information inside the truncated length, in accordance with the first probability from the first computation storage circuit, the second probability from the second computation storage circuit and the third probability from the third computation storage circuit;
• and
• an output terminal which outputs the soft output values.