Rice type fading generator class. More...

#include <itpp/comm/channel.h>

Inheritance diagram for itpp::Rice_Fading_Generator:

Public Member Functions
	Rice_Fading_Generator (double norm_doppler, DOPPLER_SPECTRUM spectrum=Jakes, int no_freq=16, RICE_METHOD method=MEDS)
	Default constructor.

virtual	~Rice_Fading_Generator ()
	Destructor.

virtual void	set_doppler_spectrum (DOPPLER_SPECTRUM spectrum)
	Set Doppler spectrum.

virtual void	set_no_frequencies (int no_freq)
	Set number of Doppler frequencies.

virtual void	set_rice_method (RICE_METHOD method)
	Set calculation method of Doppler frequencies and amplitudes.

virtual DOPPLER_SPECTRUM	get_doppler_spectrum () const
	Return Doppler spectrum.

virtual int	get_no_frequencies () const
	Get number of Doppler frequencies.

virtual RICE_METHOD	get_rice_method () const
	Get calculation method of Doppler frequencies and amplitudes.

virtual void	init ()
	Initialize the generator.

virtual void	generate (int no_samples, cvec &output)
	Generate no_samples values from the fading process.

virtual void	set_norm_doppler (double norm_doppler)
	Set normalized Doppler.

virtual void	set_LOS_doppler (double relative_doppler)
	Set relative Doppler (compared to the maximum Doppler) for the LOS component.

virtual void	set_time_offset (int offset)
	Set time offset in samples.

virtual double	get_norm_doppler () const
	Return normalized Doppler.

virtual double	get_LOS_doppler () const
	Get relative Doppler (compared to the maximum doppler) for the LOS component.

virtual double	get_time_offset () const
	Get time offset in samples.

virtual void	shift_time_offset (int no_samples)
	Shift generator time offset by a number of samples.

cvec	generate (int no_samples)
	Generate no_samples values from the fading process.

void	set_LOS_power (double relative_power)
	Set relative LOS power.

virtual void	set_filter_length (int filter_length)
	Set FIR filter length (for FIR fading generator)

double	get_LOS_power () const
	Get relative power of LOS component (Rice factor)

virtual int	get_filter_length () const
	Set FIR filter length (for FIR fading generator)

Protected Member Functions
void	init_MEDS ()
	Init function for MEDS method.

void	add_LOS (int idx, std::complex< double > &sample)
	add LOS component to the sample with index idx

Protected Attributes
DOPPLER_SPECTRUM	dopp_spectrum

int	Ni
	Number of sine waves in a Gaussian process.

RICE_METHOD	rice_method
	Rice process generation method.

double	n_dopp
	Normalized maximum Doppler frequency.

double	los_dopp
	Relative Doppler on LOS component (0.7 by default)

double	time_offset
	Time offset in samples (time state in the generator)

bool	init_flag
	signals if generator is initialized or not

double	los_power
	Relative power of LOS component compared to diffuse component (K factor)

double	los_diffuse
	Diffuse component: sqrt(1 / (1 + los_power))

double	los_direct
	Direct component: sqrt(los_power / (1 + los_power))


vec	f1

vec	f2

vec	c1

vec	c2

vec	th1

vec	th2


double	f01

double	f02

Detailed Description

Rice type fading generator class.

Author: Tony Ottosson, Adam Piatyszek and Zbigniew Dlugaszewski

A Rice generator is a generator of the form:

$\tilde \mu_i(t) = \sum_{n=1}^{N_i} c_{i,n} \cos(2\pi f_{i,n} t + \theta_{i,n})$

Here $c_{i,n}$ , $f_{i,n}$ , and $\theta_{i,n}$ are the Doppler coefficients, discrete Doppler frequencies, and Doppler phases, respectively. Rice showed that a generator of this form can perfectly model a Gaussian process when $N_i \rightarrow \infty$ . When generating a fading pattern we need a complex-valued generator

$\tilde \mu(t) = \tilde \mu_1(t) + j \tilde \mu_2(t)$

Parameters that define the generator are the normalized Doppler and the doppler spectrum. Possible values of the Doppler spectrum are:

Jakes, J, Classic, C: The classical Jakes spectrum shape. This method can be also used for modelling LOS paths, by setting the LOS relative power (Rice factor) and LOS Doppler parameters.
GaussI, GI, Gauss1, G1
GaussII, GII, Gauss2, G2

Furthermore also the number of sine waves, $ N_i $ and method used to calculate the parameters $c_{i,n}$ , $f_{i,n}$ , and $\theta_{i,n}$ can be specified. For now the only method defined for calculating the parameters is the Method of Exact Doppler Spread (MEDS). See [Pat02] for more details.