Receiving device and receiving method

A receiving method, comprising:

• a decoding procedure using a decoder which performs decoding process corresponding to a convolutional code on a received signal sequence and an error correcting part performing error correction on a decoding result sequence obtained by the decoder to generate an error corrected decoded data sequence containing error corrected decoded data corresponding to the received signal sequence;
• an iteration controlling procedure causing the decoder and the error correcting part to iteratively perform first error correct decoding on the received signal sequence;
• a penalty calculating procedure calculating, based on whether a success of the error correction is indicated or not by auxiliary information obtained from the error correcting part in second error correct decoding performed before the first error correct decoding, respective penalties for a branch transiting with a decoding result which matches the error corrected decoded data obtained in the error correcting part and a branch transiting with a decoding result which does not match the error corrected decoded data;
• and
• a penalty reflecting procedure calculating a branch metric by reflecting the penalties on a basic value so as to decrease likelihood ratio of each of the branches, and inputting the branch metric being calculated to the decoder as a feedback value from the second error correct decoding, the basic value being equivalent to a distance between each received signal contained in the received signal sequence and coder outputs in a state transition model to which a coder corresponding to the decoder complies.

A receiving device, comprising:

• a soft decision converting part converting each of components of coordinates of signal points corresponding to respective symbols included in an input received signal sequence into a sequence of soft-decision inputs represented by a predetermined number of bits in order for each of coordinate values corresponding to binary hard decision outputs to be converted to a value having a sign corresponding to the binary hard decision and having an absolute value equal to a maximum value among absolute values for conversion results;
• a decoder performing decoding process corresponding to decoding of a convolutional code on the sequence of soft-decision inputs obtained by the soft decision converting part;
• an error correcting part performing error correction on a decoding result sequence obtained by the decoder to generate an error corrected decoded data sequence containing error corrected decoded data corresponding to the received signal sequence;
• and
• an iteration controlling part causing the decoder and the error correcting part to iteratively perform first error correct decoding on the received signal sequence, wherein
• the decoder comprises a metric calculating part calculating, based on a correlation value of each of the soft-decision inputs input to the decoder and coder outputs in a state transition model to which a coder corresponding to the decoder complies, a branch metric which a larger value indicates more closeness to a maximum likelihood, and providing the branch metric being calculated to maximum likelihood estimation in the decoder;
• and
• the metric calculating part comprises:
• a weight calculating part obtaining, when error correct decoding is performed iteratively by the decoder and the error correcting part, a weight reflecting auxiliary information corresponding to each of the soft-decision inputs input to the decoder based on the auxiliary information and error corrected decoded data obtained from the error correcting part in a process of second error correct decoding performed before the first error correct decoding;
• and
• a weight adding part adding the weight calculated by the weight calculating part to the branch metric obtained for each of corresponding soft-decision inputs.