# US 2012/0254685 A1 · Readdressing decoder for quasi-cyclic low-density parity-check and method thereof

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

- Number: US 2012/0254685 A1
- Document ID: US20120254685A1
- Status: Published application
- Kind: A1
- Application number: 13/208,814
- Filed: 2011-08-12
- Published: 2012-10-04
- Earliest claimed: 2011-03-29
- Assignees: MStar Semiconductor, Inc.
- Inventors: Yu-Hsien Ku; Tung-sheng Lin; Tai-Lai Tung
- Web page: https://patentgrep.com/patent/US20120254685A1

## CPC

- H03M 13/6552: {DVB-T2}
- H03M 13/1117: {using approximations for check node processing, e.g. an outgoing message is depending on the signs and the minimum over the magnitudes of all incoming messages according to the min-sum rule}
- H03M 13/1137: {Partly parallel processing, i.e. sub-blocks or sub-groups of nodes being processed in parallel}
- H03M 13/1165: {QC-LDPC codes as defined for the digital video broadcasting [DVB] specifications, e.g. DVB-Satellite [DVB-S2]}
- 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 readdressing decoder for QC-LDPC decoding including a memory, a controller and parallel processors is provided. The memory stores a QC-LDPC matrix including sub-matrices respectively addressed with a corresponding address. The controller readdresses each of the sub-matrices into divided matrices and defines each of the divided matrices into a first address group and a second address group. The controller further respectively transmits the divided matrices of the first address group and the second address group to the parallel processors to perform correction algorithm.

## Claims

### Claim 1

A readdressing decoder for decoding a quasi-cyclic low-density parity-check (QC-LDPC) code, comprising:
a memory that stores a QC-LDPC matrix comprising a plurality of sub-matrices each having a corresponding address;
a controller that respectively readdresses the sub-matrices into a plurality of divided matrices, and defines the divided matrices into a first address group or a second address group; and
a plurality of parallel processors, coupled to the controller, that performs a correction process on the divided matrices of the first address group and the divided matrices of the second address group to generate a first correction result and a second correction result;
wherein, a ratio between a number of the divided matrices of the first address group or a number of the divided matrices of the second address group and a number of the parallel processors is an integer.

### Claim 2 (depends on 1)

The readdressing decoder according to claim 1, wherein the controller determines at least one axis index of the sub-matrices to readdress the sub-matrices.

### Claim 3 (depends on 2)

The readdressing decoder according to claim 2, wherein the first correction result and the second correction result respectively comprise an x-axis index and a y-axis index of a corresponding divided matrix.

### Claim 4 (depends on 1)

The readdressing decoder according to claim 1, wherein a ratio between the sub-matrices and the divided matrices corresponding to the sub-matrices is a first ratio, a ratio between a number of the sub-matrices and a number of the parallel processors is a second ratio, and a ratio between the first ratio and the second ratio is an integer.

### Claim 5 (depends on 1)

The readdressing decoder according to claim 1, wherein the parallel processors perform the correction process with min-sum algorithms by comparing minimum values of the first correction result and the second correction result.

### Claim 6 (depends on 3)

The readdressing decoder according to claim 3, wherein the number of the sub-matrices is 360, the number of the parallel processors is 180, and the parallel processors perform the correction process according to the y-axis indices corresponding to the divided matrices.

### Claim 7 (depends on 3)

The readdressing decoder according to claim 3, wherein the number of the sub-matrices is 360, the number of the parallel processors is 180, and the parallel processors perform the correction process according to the x-axis indices corresponding to the divided matrices.

### Claim 8 (depends on 1)

The readdressing decoder according to claim 1, wherein the controller further comprises a readdressing controller and a barrel shifter, the readdressing controller generates the first correction result and the second correction result, and the barrel shifter converts the sub-matrices to unit matrices by address shifting according to the first correction result and the second correction result.

### Claim 9 (depends on 1)

The readdressing decoder according to claim 1, wherein the memory comprises a plurality of memory units, and each of the sub-matrices is stored in one of the memory units respectively.

### Claim 10 (depends on 1)

The readdressing decoder according to claim 1, wherein when the sub-matrices are respectively a M*M matrix and a ratio between a number of the divided matrices corresponding to the first address group and a number of the parallel processors is n, a length of the divided matrices is M/n, and the controller compares the axis index with the length of the divided matrices; when the axis index is smaller than M/n, the axis index is same as the corresponding addresses of the sub-matrices of the divided matrices, and the axis index is address to a first divided matrix; and when the axis index is greater than M/n and smaller than 2M/n, the axis index is addressed to a second divided matrix, and the x-axis index is reset as x-axis index minus M/n.

### Claim 11

A readdressing method for decoding a QC-LDPC code, comprising:
readdressing a plurality of sub-matrices into a plurality of divided matrices according to a divided matrix number, respectively, and defining the divided matrices to a first address group or a second address group; and
performing a correction process on the divided matrices associated with the first address group and the divided matrices associated with the second address group.

### Claim 12 (depends on 11)

The method according to claim 11, wherein the step of respectively readdressing the sub-matrices into the plurality of divided matrices further comprises:
determining at least one axis index of the sub-matrices, and readdressing the sub-matrices according to the axis indices.

### Claim 13 (depends on 12)

The method according to claim 12, further comprising:
generating a first correction result and a second correction result, the first correction result and the second correction result respectively comprising an x-axis index and a y-axis index of the divided matrices respectively.

### Claim 14 (depends on 11)

The method according to claim 11, wherein the correction process are performed by a plurality of parallel processors; and a ratio between the sub-matrices and the divided matrices corresponding to the sub-matrices is a first ratio, a ratio between a number of the sub-matrices and a number of the parallel processors is a second ratio, and a ratio between the first ratio and the second ratio is an integer.

### Claim 15 (depends on 14)

The method according to claim 14, wherein the parallel processors perform the correction process with min-sum algorithms by comparing minimum values of the first correction result and the second correction result.

### Claim 16 (depends on 14)

The method according to claim 14, wherein the number of the sub-matrices is 360, the number of the parallel processors is 180, and the parallel processors perform the correction process according to the y-axis indices corresponding to the divided matrices.

### Claim 17 (depends on 14)

The method according to claim 14, wherein the number of the sub-matrices is 360, the number of the parallel processors is 120, and the parallel processors perform the correction process according to the x-axis indices corresponding to the divided matrices.

### Claim 18 (depends on 14)

The method according to claim 14, further comprising:
converting the sub-matrices to unit matrices by address shifting according to the first correction result and the second correction result.

### Claim 19 (depends on 11)

The method according to claim 11, wherein each of the sub-matrices is stored in one of a plurality of memory units respectively.

### Claim 20 (depends on 11)

The method according to claim 11, wherein in the step of respectively readdressing the sub-matrices into the plurality of divided matrices, when the sub-matrices are respectively a M*M matrix and a ratio between a number of the divided matrices corresponding to the first address group and a number of the parallel processors is n, a length of the divided matrices is M/n, and the controller compares the axis index with the length of the divided matrices; when the axis index is smaller than M/n, the axis index is same as the corresponding addresses of the sub-matrices of the divided matrices, and the axis index is address to a first divided matrix; and when the axis index is greater than M/n and smaller than 2M/n, the axis index is addressed to a second divided matrix, and the x-axis index is reset as x-axis index minus M/n.
