# US 2010/0169745 A1 · Soft output decoder, iterative decoder, and soft decision value calculating method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0169745 A1
- Document ID: US20100169745A1
- Status: Published application
- Kind: A1
- Application number: 12/377,139
- Filed: 2007-08-21
- Published: 2010-07-01
- Earliest claimed: 2006-08-22
- Assignees: Panasonic Corporation
- Inventors: Hiroyuki MOTOZUKA
- Web page: https://patentgrep.com/patent/US20100169745A1

## CPC

- H03M 13/2957: {Turbo codes and decoding}
- H03M 13/3905: {Maximum a posteriori probability [MAP] decoding or approximations thereof based on trellis or lattice decoding, e.g. forward-backward algorithm, log-MAP decoding, max-log-MAP decoding}
- H03M 13/3922: {Add-Compare-Select [ACS] operation in forward or backward recursions}
- H03M 13/3927: {Log-Likelihood Ratio [LLR] computation by combination of forward and backward metrics into LLRs}
- H03M 13/3972: {using sliding window techniques or parallel windows}
- H03M 13/6502: {Reduction of hardware complexity or efficient processing}
- H03M: CODING; DECODING; CODE CONVERSION IN GENERAL (using fluidic means F15C4/00; optical analogue/digital converters G02F7/00; coding, decoding or code conversion, specially adapted for particular applications, see the relevant subclasses, e.g. G01D, G01R, G06F, G06T, G09G, G10L, G11B, G11C, H04B, H04L, H04M, H04N; ciphering or deciphering for cryptography or other purposes involving the need for secrecy G09C)

## Abstract

A soft output decoder capable of performing decoding with a small computational complexity necessary for likelihood calculation thereby to reduce the scale of the operation circuit and to shorten the processing delay time. The soft output decoder (100) comprises a &ggr;α calculating section (101), timing adjusters (102, 103), a &ggr;β calculation section (104), and a &Lgr; calculating unit (110). The &Lgr; calculating unit (110) is composed of a hard decision decoding section (111), a loss likelihood calculating section (112), a win likelihood storage section (113), and a subtractor (114). The likelihoods determined by a Max-log-MAP decoder are separated into win likelihoods and loss likelihoods, and only the loss likelihoods are calculated. A hard decision decoding section (111) for sequentially specifying the states through which the maximum likelihood path passes is used for a method for distinguishing the win and loss likelihoods.

## Claims

### Claim 1

A soft output decoder that receives code words corresponding to a plurality of information words and calculates soft decision decoding results of the information words from a plurality of likelihood values corresponding to a plurality of candidate values adopted by the information words, the soft output decoder comprising:
a hard decision decoding section that decides one of the plurality of candidate values as an estimated value; and
a likelihood calculating section that bans calculation of a likelihood value corresponding to the estimated value and calculates likelihood values corresponding to other candidate values in the plurality of likelihood values.

### Claim 2 (depends on 1)

The soft output decoder according to claim 1, wherein:
the hard decision decoding section outputs α value 0 or 1 as the estimated value; and
the likelihood calculating section bans calculation of a likelihood value corresponding to the estimated value.

### Claim 3 (depends on 1)

The soft output decoder according to claim 1, wherein the likelihood calculating section calculates one likelihood value having a smaller likelihood when a number of the plurality of candidates is two.

### Claim 4 (depends on 1)

The soft output decoder according to claim 1, wherein, when a number of the plurality of candidate values is four, the likelihood calculating section excludes one maximum likelihood and calculates a rest of three likelihood values.

### Claim 5 (depends on 1)

The soft output decoder according to claim 1, further comprising a soft decision value calculating section that, when the likelihood value corresponding to the estimated value is a winning likelihood, calculates a difference between the winning likelihood and one of outputs of the likelihood calculating section.

### Claim 6 (depends on 5)

The soft output decoder according to claim 5, wherein the winning likelihood is a prescribed fixed value.

### Claim 7 (depends on 1)

The soft output decoder according to claim 1, further comprising:
a boundary winning likelihood calculating section that defines one of the information words as a boundary portion and calculates a winning likelihood in the boundary portion; and
a storing section that holds the boundary winning likelihood,
wherein the soft decision calculating section calculates a soft decision decoding result for information words other than the boundary portion using the boundary winning likelihood.

### Claim 8 (depends on 1)

The soft output decoder according to claim 1, wherein:
the information words adopt α value 0 or 1; and
the soft decision decoding results of the information words are outputted based on a difference between a likelihood value corresponding to α value 0 and a likelihood value corresponding to value 1.

### Claim 9 (depends on 1)

The soft output decoder according to claim 1, wherein the plurality of candidate values comprise two candidate values 0 and 1.

### Claim 10 (depends on 1)

The soft output decoder according to claim 1, wherein the plurality of candidate values comprise four candidate values of 00, 01, 10 and 11.

### Claim 11 (depends on 5)

The soft output decoder according to claim 5, wherein:
the received cord word is divided into a plurality of blocks;
soft decision is carried out per block; and
a winning likelihood varying between blocks is used.

### Claim 12 (depends on 1)

The soft output decoder according to claim 1, wherein likelihood calculation is carried out using a sliding window for data subjected to error correction coding by turbo code.

### Claim 13

The soft output decoder comprising:
a boundary winning likelihood calculating section that defines one of both ends of the sliding window as a boundary portion and calculates a winning likelihood in the boundary portion; and
a storing section that holds the boundary winning likelihood,
wherein the soft decision calculating section calculates soft decision decoding results for information words other than the boundary portion using the boundary winning likelihood.

### Claim 14 (depends on 1)

The soft output decoder according to claim 1, further comprising:
a boundary likelihood calculating section that defines one of the information words as a boundary portion and finds a plurality of boundary likelihood values corresponding to the plurality candidate values adopted by the information words in the boundary portion; and
an initial state number specifying section that specifies a state number to which a maximum likelihood branch is connected using the boundary likelihoods,
wherein the hard decision decoding section decides an estimated value for information words other than the boundary portion using the state number.

### Claim 15 (depends on 14)

The soft output decoder according to claim 14, wherein, when receiving an output stop command for a soft decision value and when receiving an output command for a hard decision value, the likelihood calculating section bans likelihood calculation for the information words other than the boundary portion.

### Claim 16 (depends on 14)

An iterative decoding apparatus that repeats soft output decoding, the iterative decoding apparatus comprising:
a soft output decoder according to claim 14; and
a commanding section that, when repetition is final, commands the soft output decoder to output the hard decision value and stop outputting the soft decision value.

### Claim 17

A soft decision value calculating method for receiving code words corresponding to a plurality of information words and calculating soft decision decoding results of the information words from a plurality of likelihood values corresponding to a plurality of candidate values adopted by the information words, the soft decision value calculating method comprising:
deciding one of the plurality of candidate values as an estimated value; and
banning calculation of a likelihood value corresponding to the estimated value and calculating likelihood values corresponding to other candidate values in the plurality of likelihood values.