3.2.4.3.1. LDPC Encoder parameters

3.2.4.3.1.1. --enc-cw-size, -N Required

Type:integer
Examples:--enc-cw-size 1024

Set the codeword size \(N\).

Note

This parameter value is automatically deduced if the \(H\) parity matrix is given with the --dec-h-path parameter or if the \(G\) generator matrix is given with the --enc-g-path parameter.

3.2.4.3.1.2. --enc-info-bits, -K Required

Type:integer
Examples:--enc-info-bits 512

Set the number of information bits \(K\).

Note

This parameter value is automatically deduced if the \(G\) generator matrix is given with the --enc-g-path parameter.

Note

In some cases, this parameter value can be automatically deduced if the \(H\) parity matrix is given with the --dec-h-path parameter. For regular matrices, \(K = N - M\) where \(N\) and \(M\) are the \(H\) parity matrix dimensions. For non-regular matrices, \(K\) has to be given.

3.2.4.3.1.3. --enc-type

Type:text
Allowed values:LDPC LDPC_H LDPC_DVBS2 LDPC_IRA LDPC_QC AZCW COSET USER
Default:AZCW
Examples:--enc-type AZCW

Type of the encoder to use in the simulation.

Description of the allowed values:

Value Description
LDPC Select the generic encoder that encode from a given \(G\) generator matrix (to use with the --enc-g-path parameter).
LDPC_H Build the \(G\) generator matrix from the given \(H\) parity matrix and then encode with the LDPC method (to use with the --dec-h-path parameter).
LDPC_DVBS2 Select the optimized encoding process for the DVB-S2 \(H\) matrices (to use with the --enc-cw-size, -N and --enc-info-bits, -K parameters).
LDPC_IRA Select the optimized encoding process for the IRA \(H\) parity matrices (to use with the --dec-h-path parameter).
LDPC_QC Select the optimized encoding process for the QC \(H\) parity matrices (to use with the --dec-h-path parameter).
AZCW See the common --enc-type parameter.
COSET See the common --enc-type parameter.
USER See the common --enc-type parameter.

Note

The LDPC_DVBS2 encoder type allow the simulation of the DVB-S2 standard but without the BCH code. All matrices described by the standard (Tables 5a/5b page 22-23) are available. You just need to give to the arguments --enc-info-bits, -K and --enc-cw-size, -N the real \(K\) and \(N\) LDPC dimensions, respectively.

3.2.4.3.1.4. --enc-g-path

Type:file
Rights:read only
Examples:--enc-g-path example/path/to/the/G_matrix.alist

Give the path to the \(G\) generator matrix in an AList or QC formated file.

3.2.4.3.1.5. --enc-g-method

Type:text
Allowed values:IDENTITY LU_DEC
Default:IDENTITY
Examples:--enc-g-method IDENTITY

Specify the method used to build the \(G\) generator matrix from the \(H\) parity matrix when using the LDPC_H encoder.

Description of the allowed values:

Value Description
IDENTITY Generate an identity on \(H\) to get the parity part.
LU_DEC Generate a hollow \(G\) thanks to the LU decomposition with a guarantee to have the systematic identity. Do not work with irregular matrices.

LU_DEC method is faster than IDENTITY.

3.2.4.3.1.6. --enc-g-save-path

Type:file
Rights:write only
Examples:--enc-g-save-path example/path/to/the/generated/G_matrix.alist

Set the file path where the \(G\) generator matrix will be saved (AList file format). To use with the LDPC_H encoder.

Hint

When running the LDPC_H encoder, the generation of the \(G\) matrix can take a non-negligible part of the simulation time. With this option the \(G\) matrix can be saved once for all and used in the standard LDPC decoder after.

Warning

This option is not thread-safe, please run it on a single thread with the --sim-threads, -t parameter.