gr::dvbs2rx::bch_codec< T, P > Class Template Reference

BCH coder/decoder. More...

#include <bch.h>

Public Member Functions
	bch_codec (const galois_field< T > *const gf, uint8_t t, uint32_t n=0)
	Construct a new BCH coder/decoder object. More...
T	encode (const T &msg) const
	Encode an input message. More...
void	encode (u8_cptr_t msg, u8_ptr_t codeword) const
std::vector< T >	syndrome (const T &codeword) const
	Compute the syndrome of a received codeword. More...
std::vector< T >	syndrome (u8_cptr_t codeword) const
gf2m_poly< T >	err_loc_polynomial (const std::vector< T > &syndrome) const
	Compute the error-location polynomial. More...
std::vector< T >	err_loc_numbers (const gf2m_poly< T > &sigma) const
	Compute the error-location numbers. More...
T	decode (T codeword) const
	Decode an input codeword. More...
int	decode (u8_cptr_t codeword, u8_ptr_t decoded_msg) const
const gf2_poly< P > &	get_gen_poly () const
	Get the generator polynomial object. More...
uint16_t	get_n () const
	Get the codeword length n. More...
uint16_t	get_k () const
	Get the message length k. More...

Detailed Description

template<typename T, typename P>
class gr::dvbs2rx::bch_codec< T, P >

BCH coder/decoder.

Template Parameters

T	Base type for the Galois Field elements.
P	Base type for the GF(2) generator polynomial.

Constructor & Destructor Documentation

bch_codec()

template<typename T , typename P >

gr::dvbs2rx::bch_codec< T, P >::bch_codec	(	const galois_field< T > *const	gf,
		uint8_t	t,
		uint32_t	n = 0
	)

Construct a new BCH coder/decoder object.

Parameters

gf	Reference Galois field.
t	Target error correction capability.
n	Target codeword length in bits.

Note

The default codeword length is n=(2^m - 1), where m is the dimension of the GF(2^m) Galois Field, so the codeword length is inferred from the gf parameter. Alternatively, a codeword length lower than (2^m - 1) and greater than the generator polynomial's degree can be specified through parameter n. In this case, the constructor attempts to construct a shortened (n - s, k - s) BCH code, with parameter s equal to (2^m - 1) - n.

Member Function Documentation

decode() [1/2]

template<typename T , typename P >

T gr::dvbs2rx::bch_codec< T, P >::decode

(

T

codeword

)

const

Decode an input codeword.

Parameters

codeword

n-bit input codeword.

Returns

T Decoded message.

decode() [2/2]

template<typename T , typename P >

int gr::dvbs2rx::bch_codec< T, P >::decode	(	u8_cptr_t	codeword,
		u8_ptr_t	decoded_msg
	)		const

This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

Parameters

codeword	Pointer to the received codeword with n/8 bytes.
decoded_msg	Pointer to the decoded message buffer with space for k/8 bytes.

Returns

int Number of bit errors corrected by the decoder. Set to 0 when the message is error-free and -1 when the decoding fails to correct all errors such that the decoded message has residual bit errors.

Note

The caller should make sure the pointers point to memory regions with enough data and space.

This bytes-based decoding is only supported when n and k are multiples of 8.

encode() [1/2]

template<typename T , typename P >

T gr::dvbs2rx::bch_codec< T, P >::encode

(

const T &

msg

)

const

Encode an input message.

Parameters

msg	k-bit input message.

Returns

T Resulting codeword.

Note

The parity digits are placed at the least significant (lower order) n - k bit positions, whereas the original (systematic) message is placed at the most significant k bit positions, from bit n-1 to bit n-k.

Only use this implementation if the message type T can hold n bits. Otherwise, use the implementation based on an array of bytes.

encode() [2/2]

template<typename T , typename P >

void gr::dvbs2rx::bch_codec< T, P >::encode	(	u8_cptr_t	msg,
		u8_ptr_t	codeword
	)		const

This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

Parameters

msg	Pointer to the input message with k/8 bytes.
codeword	Pointer to the codeword buffer with space for n/8 bytes.

Note

Since the code is systematic, the first k/8 bytes of the resulting codeword hold the original message, whereas the remaining bytes contain the parity digits. A serializer should read the codeword from the first byte to the last byte to preserve the order.

The caller should make sure the pointers point to memory regions with enough data and space.

This bytes-based encoding is only supported when n and k are multiples of 8.

err_loc_numbers()

template<typename T , typename P >

std::vector< T > gr::dvbs2rx::bch_codec< T, P >::err_loc_numbers

(

const gf2m_poly< T > &

sigma

)

const

Compute the error-location numbers.

The error-location numbers are the numbers from GF(2^m) corresponding to the reciprocal of the roots of the error-location polynomial. An error-location number alpha^j indicates there is an error in the j-th bit of the received codeword.

Parameters

sigma

Error location polynomial.

Returns

std::vector<T> Error location numbers, a vector of elements from GF(2^m).

err_loc_polynomial()

template<typename T , typename P >

gf2m_poly< T > gr::dvbs2rx::bch_codec< T, P >::err_loc_polynomial

(

const std::vector< T > &

syndrome

)

const

Compute the error-location polynomial.

The error-location polynomial is a polynomial over GF(2^m) whose roots indicate the location of bit errors. The implementation for finding this polynomial is based on the simplified Berlekamp's iterative algorithm, which works for binary BCH codes.

Parameters

syndrome

Syndrome vector with 2t elements.

Returns

gf2m_poly<T> Error-location polynomial, a polynomial over GF(2^m).

get_gen_poly()

template<typename T , typename P >

const gf2_poly< P > & gr::dvbs2rx::bch_codec< T, P >::get_gen_poly ( ) const

inline

Get the generator polynomial object.

Returns

const gf2_poly

& Generator polynomial.

get_k()

template<typename T , typename P >

uint16_t gr::dvbs2rx::bch_codec< T, P >::get_k ( ) const

inline

Get the message length k.

Returns

uint16_t Message length.

get_n()

template<typename T , typename P >

uint16_t gr::dvbs2rx::bch_codec< T, P >::get_n ( ) const

inline

Get the codeword length n.

Returns

uint16_t Codeword length.

syndrome() [1/2]

template<typename T , typename P >

std::vector< T > gr::dvbs2rx::bch_codec< T, P >::syndrome

(

const T &

codeword

)

const

Compute the syndrome of a received codeword.

Parameters

codeword

Received codeword.

Returns

std::vector<T> Syndrome as a vector with 2t GF(2^m) elements or an empty vector when the codeword is error-free.

Note

An empty vector is returned for an error-free codeword for CPU efficiency. The caller should check the size of the returned vector to determine whether the codeword is error-free or not.

syndrome() [2/2]

template<typename T , typename P >

std::vector< T > gr::dvbs2rx::bch_codec< T, P >::syndrome

(

u8_cptr_t

codeword

)

const

This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

Parameters

codeword

Pointer to u8 array with the received codeword.

---

The documentation for this class was generated from the following file:

• bch.h