This class implements "coding and modulation" using an LDCP codec through the functions modulate and demodulate. The function modulate accepts a vector of bits and generates a vector of complex symbols (e.g. 16QAM symbols). The funcion demodulate conversally takes a vector of noise symbols and calculates the maximum aposteriori estimates of the transmitted bits. More...

#include <AMC.hpp>

Public Member Functions
	AMC (void)
	Constructor without arguments. Call init after construction.
void	init (uint32_t code_word_size, bool load_from_file)
	Initialize an object of type AMC.
void	modulate (uint32_t modulation_order, uint32_t codec_ix, bvec message, bvec &transmitted, cvec &symbols, int skip_symbols)
	Perform modulation.
void	encode (uint32_t codec_ix, bvec message, bvec &transmitted)
	Identical to modulation except fewer inputs and outputs.
int	codec_message_size (uint32_t codec_ix)
	Return code message size i.e. number payload bits per codeword.
int	no_of_codecs (void)
	Number of codecs. Currently 4.
double	evm (uint32_t modulation_order, bvec &transmitted, cvec &symbols, int skip_symbols)
	Calculate error vector magnitude.
void	demodulate (const cvec &input_symbols, double noise_variance, uint32_t modulation_order, uint32_t codec_ix, bvec &message_hat, bvec &transmitted_hat)
	Demodulate a vector of noisy complex symbols (e.g 16QAM).
double	noise_variance (const cvec &symbols, uint32_t modulation_order)
	Estimate noise variance in a set of symbols with known modulation_order.
void	randb_thread_safe (int size, bvec &out, uint64_t *seed)
	As the name suggests, this is a thread safe version of the randb(int size, bvec &out) function in IT++. This function has an additional argument, namely the seed. This argument makes it simple to handle multiple random bit streams.
void	randb_thread_safe (int start_ix, int length, bvec &out, uint64_t *seed)
	Randomizes the bits with index start_ix to start_ix+length in the vector out.
void	run_through_all_mcs (void)
	This function simulates all combinations of coding and modulation parameters for the chose code_word_size and saves the results in the file results.it. The results can be then be read and displayed by the matlab script { plot}.
Public Attributes
bool	only_raw_bits

Detailed Description

This class implements "coding and modulation" using an LDCP codec through the functions modulate and demodulate. The function modulate accepts a vector of bits and generates a vector of complex symbols (e.g. 16QAM symbols). The funcion demodulate conversally takes a vector of noise symbols and calculates the maximum aposteriori estimates of the transmitted bits.

Definition at line 33 of file AMC.hpp.

Member Function Documentation

void AMC::demodulate	(	const cvec &	input_symbols,
		double	noise_variance,
		uint32_t	modulation_order,
		uint32_t	codec_ix,
		bvec &	message_hat,
		bvec &	transmitted_hat
	)

Demodulate a vector of noisy complex symbols (e.g 16QAM).

Parameters:

input_symbols	Demodulate the symbols in positions skip_symbols,skip_symbols+1,...,N+skip_symbols-1, where N=code_word_size/log2(modulation_order). The parameter skip_symbols is by default zero but can be changed.
noise_variance	The variance of the noise in the input symbols.
modulation_order	Modulation order 4, 16, 64 or 256 (QAM).
codec_ix	Index of the codec 0,1 and 2, correspond to rate 1/4, 1/2 and 3/4, respectively.
message_hat	The bit estimates according to the iterative MAP decoder. If only_raw_bits is true, then these estimates are actually not calculated. This is useful to reduce computation time. If the raw symbols are saved of if all data is logged (see save_raw_input_on_file of four_multi) it is often possible to do the decoding offline after the experiements. The boolean only_raw_bits is true by default.
transmitted_hat	The hard bit estimates.

Definition at line 103 of file AMC.cpp.

double AMC::evm	(	uint32_t	modulation_order,
		bvec &	transmitted,
		cvec &	symbols,
		int	skip_symbols
	)

Calculate error vector magnitude.

Parameters:

modulation_order. Modulation order 4, 16, 64 or 256 (QAM).

Definition at line 78 of file AMC.cpp.

void AMC::init	(	uint32_t	code_word_size,
		bool	load_from_file
	)

Initialize an object of type AMC.

Parameters:

code_word_size	Three codes are generated with rate 1/4, 1/2, 3/4 and 5/6. This parameter provides the code word size.
load_from_file.	If load_from_file=true then the codecs are read from code_data1.it, code_data2.it and code_data3.it, otherwise the codecs are generated. When code_word_size is changed, the codes need to be re-generated. The user is responsible for keeping track of this.

Definition at line 33 of file AMC.cpp.

void AMC::modulate	(	uint32_t	modulation_order,
		uint32_t	codec_ix,
		bvec	message,
		bvec &	transmitted,
		cvec &	symbols,
		int	skip_symbols
	)

Perform modulation.

Parameters:

modulation_order	Valid values are 4, 16, 64 and 256. QAM is used.
code_ix	Index of the codec 0,1 and 2, correspond to rate 1/4, 1/2 and 3/4, respectively.
message	Input bit vector of length code_word_size*rate or code_message_size, see below.
transmitted	Generated encoded bit vector of length code_word_size.
symbols.	Generated vector of complex symbols of length N=code_word_size/log2(modulation_order). The generated symbols are written into positions skip_symbols,skip_symbols+1, ...,N+skip_symbols-1.
skip_symbols.	See symbols.

Definition at line 54 of file AMC.cpp.

double AMC::noise_variance	(	const cvec &	symbols,
		uint32_t	modulation_order
	)

Estimate noise variance in a set of symbols with known modulation_order.

Returns:: Estimated noise_variance.

Parameters:

symbols	Vector of modulation symbols.
modulation_order	Modulation order of QAM, 4 16, 64 or 256.

Definition at line 121 of file AMC.cpp.

void AMC::randb_thread_safe	(	int	size,
		bvec &	out,
		uint64_t *	seed
	)

As the name suggests, this is a thread safe version of the randb(int size, bvec &out) function in IT++. This function has an additional argument, namely the seed. This argument makes it simple to handle multiple random bit streams.

Parameters: