In several places we have a real valued signal x(t) that needs to be converted to an analytic signal g(t) such that real(g(t)) == x(t) and imag(g(t)) = shift_by_90degrees(x(t));. More...

#include <HilbertTransformer.hxx>

Inheritance diagram for SoDa::HilbertTransformer:

Public Member Functions
	HilbertTransformer (unsigned int inout_buffer_length, unsigned int filter_length=256)
	constructor – build a Hilbert Transformer More...

unsigned int	applyIQ (std::complex< float > inbuf, std::complex< float > outbuf, float gain=1.0)
	Perform a hilbert transform on the QUADRATURE signal in the input buffer. More...

unsigned int	apply (std::complex< float > inbuf, std::complex< float > outbuf, bool pos_sided=true, float gain=1.0)
	Perform a hilbert transform on the INPHASE signal in the input buffer. More...

unsigned int	apply (float inbuf, std::complex< float > outbuf, bool pos_sided=true, float gain=1.0)
	Perform a hilbert transform on the signal in the floating point input buffer. More...

std::ostream &	dump (std::ostream &os)

Public Member Functions inherited from SoDa::SoDaBase
	SoDaBase (const std::string &oname)
	The constructor – pass a name for the object. More...

std::string &	getObjName ()
	get the name of this object More...

SoDaBase *	findSoDaObject (const std::string &oname)
	find a SoDa Object by name. More...

double	getTime ()
	Get a time stamp in nS resolution that monotonically increases and that is very inexpensive (typically < 100nS). More...

Private Attributes
unsigned int	M
	these are the salient dimensions for this Overlap/Save widget (for terminology, see Lyons pages 719ff More...

unsigned int	Q
	the filter length More...

unsigned int	N
	the total length of the transform N > (M + Q-1) More...

std::complex< float > *	fft_I_input

std::complex< float > *	fft_Q_input

std::complex< float > *	fft_I_output

std::complex< float > *	fft_Q_output

std::complex< float > *	ifft_I_input

std::complex< float > *	ifft_Q_input

std::complex< float > *	ifft_I_output

std::complex< float > *	ifft_Q_output

fftwf_plan	forward_I_plan

fftwf_plan	forward_Q_plan

fftwf_plan	backward_I_plan

fftwf_plan	backward_Q_plan

std::complex< float > *	HTu_filter
	The DFT image of the hilbert transform – upper sideband. More...

std::complex< float > *	HTl_filter
	The DFT image of the hilbert transform – lower sideband. More...

std::complex< float > *	Pass_U_filter
	The DFT image of a Q/2 delay transform – used in USB. More...

std::complex< float > *	Pass_L_filter
	The DFT image of a Q/2 delay transform – used in LSB. More...

float	passthrough_gain
	the gain of the direct passthrough path. More...

float	H_transform_gain
	the gain of the Hilbert Transform path More...

Detailed Description

In several places we have a real valued signal x(t) that needs to be converted to an analytic signal g(t) such that real(g(t)) == x(t) and imag(g(t)) = shift_by_90degrees(x(t));.

HilbertTransformer provides the apply functions to convert x(t) as a float or complex<float>. Hilbert Transformer also provides a function (applyIQ) to convert a complex x(t) into g(t) such that real(g(t)) = real(x(t + tau)) and imag(g(t)) = shift_by_90deg(imag(x(t + tau)))

Definition at line 54 of file HilbertTransformer.hxx.

Constructor & Destructor Documentation

◆ HilbertTransformer()

SoDa::HilbertTransformer::HilbertTransformer	(	unsigned int	inout_buffer_length,
		unsigned int	filter_length = `256`
	)

constructor – build a Hilbert Transformer

Parameters

inout_buffer_length	the length of the input and output buffers
filter_length	the minimum length of the hilbert transform impulse response

Definition at line 52 of file HilbertTransformer.cxx.

References backward_I_plan, backward_Q_plan, fft_I_input, fft_I_output, fft_Q_input, fft_Q_output, forward_I_plan, forward_Q_plan, H_transform_gain, HTl_filter, HTu_filter, ifft_I_input, ifft_I_output, ifft_Q_input, ifft_Q_output, M, N, Pass_L_filter, Pass_U_filter, passthrough_gain, and Q.

Member Function Documentation

◆ apply() [1/2]

unsigned int SoDa::HilbertTransformer::apply	(	std::complex< float > *	inbuf,
		std::complex< float > *	outbuf,
		bool	pos_sided = `true`,
		float	gain = `1.0`
	)

Perform a hilbert transform on the INPHASE signal in the input buffer.

It is assumed that the QUADRATURE signal is zero, if not, the result is broken.

Parameters

inbuf	complex input buffer of I (real) and Q (imag) samples
outbuf	complex output buffer real is input.real delayed, and imag is Hilbert(input.real)
pos_sided	if true, swap I and Q outputs
gain	factor to apply to output buffer.

Returns: M – length of input buffer.

Definition at line 205 of file HilbertTransformer.cxx.

References backward_I_plan, backward_Q_plan, dbgctr, fft_I_input, fft_I_output, forward_I_plan, H_transform_gain, HTl_filter, HTu_filter, ifft_I_input, ifft_I_output, ifft_Q_input, ifft_Q_output, M, N, Pass_L_filter, Pass_U_filter, passthrough_gain, and Q.

Referenced by apply(), doPMTest(), doSSBTest(), dumpTest(), and SoDa::BaseBandTX::modulateAM().

◆ apply() [2/2]

unsigned int SoDa::HilbertTransformer::apply	(	float *	inbuf,
		std::complex< float > *	outbuf,
		bool	pos_sided = `true`,
		float	gain = `1.0`
	)

Perform a hilbert transform on the signal in the floating point input buffer.

Parameters

inbuf	input buffer of real (float) samples
outbuf	complex output buffer real is input delayed, and imag is Hilbert(input)
pos_sided	if true, swap I and Q outputs
gain	factor to apply to output buffer.

Returns: M – length of input buffer.

Definition at line 259 of file HilbertTransformer.cxx.

References apply(), and M.

◆ applyIQ()

unsigned int SoDa::HilbertTransformer::applyIQ	(	std::complex< float > *	inbuf,
		std::complex< float > *	outbuf,
		float	gain = `1.0`
	)

Perform a hilbert transform on the QUADRATURE signal in the input buffer.

Pass the Inphase signal through a delay filter that matches the hilbert transform

Parameters

inbuf	complex input buffer of I (real) and Q (imag) samples
outbuf	complex output buffer real is input.real delayed, and imag is Hilbert(input.imag)
gain	factor to apply to output buffer.