dsp.ComplexBandpassDecimator

Extract a frequency subband using a one-sided (complex) bandpass decimator

Description

The dsp.ComplexBandpassDecimator System object™ extracts a specific sub-band of frequencies using a one-sided, multistage, complex bandpass decimator. The object determines the bandwidth of interest using the specified CenterFrequency, DecimationFactor and Bandwidth values.

To extract a frequency subband using a complex bandpass decimator:

1. Create the dsp.ComplexBandpassDecimator object and set its properties.

2. Call the object with arguments, as if it were a function.

To learn more about how System objects work, see What Are System Objects?

This object supports C/C++ code generation and SIMD code generation under certain conditions. For more information, see Code Generation.

Creation

Syntax

bpdecim = dsp.ComplexBandpassDecimator

bpdecim = dsp.ComplexBandpassDecimator(d)

bpdecim = dsp.ComplexBandpassDecimator(d,Fc)

bpdecim = dsp.ComplexBandpassDecimator(d,Fc,Fs)

bpdecim = dsp.ComplexBandpassDecimator(Name=Value)

Description

bpdecim = dsp.ComplexBandpassDecimator creates a System object that filters each channel of the input over time using a one-sided, multistage, complex bandpass decimation filter. The object determines the bandwidth of interest using the default center frequency, decimation factor, and bandwidth values.

bpdecim = dsp.ComplexBandpassDecimator(d) creates a complex bandpass decimator object with the DecimationFactor property set to d.

bpdecim = dsp.ComplexBandpassDecimator(d,Fc) creates a complex bandpass decimator object with the DecimationFactor property set to d, and the CenterFrequency property set to Fc.

bpdecim = dsp.ComplexBandpassDecimator(d,Fc,Fs) creates a complex bandpass decimator object with the DecimationFactor property set to d, the CenterFrequency property set to Fc, and the SampleRate property set to Fs.

Example: dsp.ComplexBandpassDecimator(48e3/1e3,2e3,48e3);

bpdecim = dsp.ComplexBandpassDecimator(Name=Value) creates a complex bandpass decimator object with each specified property set to the specified value. You can use this syntax with any previous input argument combinations.

Example: dsp.ComplexBandpassDecimator(48e3/1e3,2e3,48e3,CenterFrequency=1e3);

Properties

Unless otherwise indicated, properties are nontunable, which means you cannot change their values after calling the object. Objects lock when you call them, and the release function unlocks them.

If a property is tunable, you can change its value at any time.

For more information on changing property values, see System Design in MATLAB Using System Objects.

CenterFrequency — Center frequency

0 (default) | real scalar

Center frequency of the desired band F_c, specified as a real scalar in Hz or in normalized frequency units (since R2024b).

If you set NormalizedFrequency to true, the center frequency value must be in the range (−1,1). If you set NormalizedFrequency to false, the center frequency value must be in the range (–Fs/2,Fs/2), where Fs is the value of the SampleRate property.

Tunable: Yes

Data Types: single | double

Specification — Filter design parameters

'Decimation factor' (default) | 'Bandwidth' | 'Decimation factor and bandwidth'

Filter design parameters, specified as:

• 'Decimation factor' –– The object specifies the decimation factor through the Decimation Factor property. The bandwidth of interest (BW) is computed using the following equation:

  $$BW=Fs/D$$

  where

  • Fs –– Sample rate specified through the SampleRate property.

  • D –– Decimation factor.

• 'Bandwidth' –– The object specifies the bandwidth through the Bandwidth property. The decimation factor (D) is computed using the following equation:

  $$D=\text{floor}\left(\frac{Fs}{BW+TW}\right)$$

  where

  • Fs –– Sample rate specified through SampleRate property.

  • BW –– Bandwidth of interest.

  • TW –– Transition width specified through the TransitionWidth property.

• 'Decimation factor and bandwidth' –– The decimation factor and the bandwidth of interest are specified through the DecimationFactor and Bandwidth properties.

DecimationFactor — Decimation factor

2 (default) | positive integer

Factor by which to reduce the bandwidth of the input signal, specified as a positive integer. The frame size (number of rows) of the input signal must be a multiple of the decimation factor.

Dependencies

To enable this property, set Specification to 'Decimation factor' or 'Decimation factor and bandwidth'.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

StopbandAttenuation — Stopband attenuation in dB

80 (default) | positive scalar

Stopband attenuation of the filter in dB, specified as finite positive scalar.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

TransitionWidth — Transition width

100 (default) | positive scalar

Transition width of the filter TW, specified as a positive scalar in Hz or in normalized frequency units (since R2024b).

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

Bandwidth — Width of frequency band of interest

5000 (default) | real positive scalar

Width of the frequency band of interest BW, specified as a real positive scalar in Hz or in normalized frequency units (since R2024b).

Dependencies

To enable this property, set Specification to 'Bandwidth' or 'Decimation factor and bandwidth'.

Data Types: single | double

PassbandRipple — Passband ripple (dB)

1 (default) | positive scalar

Passband ripple of the filter, specified as a positive scalar in dB.

Dependencies

To enable this property, set Specification to 'Bandwidth' or 'Decimation factor and bandwidth'.

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

MinimizeComplexCoefficients — Flag to minimize number of complex coefficients

true (default) | false

Flag to minimize the number of complex filter coefficients, specified as:

• true –– The first stage of the multistage filter is bandpass (complex coefficients) centered at the specified center frequency. The first stage is followed by a mixing stage that heterodynes the signal to DC. The remaining filter stages, all with real coefficients, follow.

• false –– The input signal is first passed through the different stages of the multistage filter. All stages are bandpass (complex coefficients). The signal is then heterodyned to DC if MixToBaseband is true, and the frequency offset resulting from the decimation is nonzero.

MixToBaseband — Flag to mix signal to baseband

true (default) | false

Flag to mix the signal to baseband, specified as:

• true –– The object heterodynes the filtered, decimated signal to DC. This mixing stage runs at the output sample rate of the filter.

• false –– The object skips the mixing stage.

Dependencies

This property applies when you set MinimizeComplexCoefficients to false.

NormalizedFrequency — Option to set frequencies in normalized units

false (default) | true

Since R2024b

Option to set frequencies in normalized units, specified as one of these values:

• true –– The center frequency, transition width, and bandwidth must be in the normalized frequency units (0 to 1).

  When you set the NormalizedFrequency property to true while creating the object and you do not set the frequency specifications, the object automatically sets the default values to normalized frequency units using the default sample rate of 44.1 kHz.

  cbd = dsp.ComplexBandpassDecimator(NormalizedFrequency=true)
  

  cbd = 
    dsp.ComplexBandpassDecimator with properties:
  
                  CenterFrequency: 0
                    Specification: 'Decimation factor'
                 DecimationFactor: 2
              StopbandAttenuation: 80
                  TransitionWidth: 0.0045
      MinimizeComplexCoefficients: true
              NormalizedFrequency: true
  

  When you set the NormalizedFrequency property to true after you create the object, you must specify the center frequency, transition width, and bandwidth in normalized units before you run the object algorithm.

  cbd = dsp.ComplexBandpassDecimator

  cbd = 
    dsp.ComplexBandpassDecimator with properties:
  
                  CenterFrequency: 0
                    Specification: 'Decimation factor'
                 DecimationFactor: 2
              StopbandAttenuation: 80
                  TransitionWidth: 100
      MinimizeComplexCoefficients: true
              NormalizedFrequency: false
                       SampleRate: 44100
  

  To specify the normalized frequency values, set NormalizedFrequency to true and manually convert the frequency values in Hz to the normalized values using the input sample rate in Hz and the interpolation factor. For example, if the input sample rate Fs is 44.1 kHz, the corresponding center frequency in normalized units is CF_Hz/(Fs/2), the corresponding bandwidth in normalized units is BW_Hz/(Fs/2), and the corresponding transition width in normalized units is TW_Hz/(Fs/2).

  cbd = dsp.ComplexBandpassDecimator;
  cbd.NormalizedFrequency = true;
  cbd.TransitionWidth = 100/(44100/2)
  

  cbd = 
    dsp.ComplexBandpassDecimator with properties:
  
                  CenterFrequency: 0
                    Specification: 'Decimation factor'
                 DecimationFactor: 2
              StopbandAttenuation: 80
                  TransitionWidth: 0.0045
      MinimizeComplexCoefficients: true
              NormalizedFrequency: true
  

• false –– The center frequency, bandwidth, and transition width values are in Hz. You can specify the input sample rate through the SampleRate property.

Data Types: logical

SampleRate — Input sample rate in Hz

44100 (default) | real positive scalar

Sampling rate of the input signal Fs in Hz, specified as a real positive scalar.

Dependencies

To enable this property, set NormalizedFrequency to false. (since R2024b)

Data Types: single | double

Usage

Syntax

y = bpdecim(x)

Description

y = bpdecim(x) filters the real or complex input signal, x, to produce the output, y. The output contains the subband of frequencies specified by the System object properties. The System object filters each channel of the input signal independently over time. The frame size (first dimension) of x must be a multiple of the decimation factor.

Input Arguments

x — Data input

vector | matrix

Data input, specified as a vector or a matrix of size P-by-Q. The number of input rows P can be arbitrary and does not have to be a multiple of the decimation factor.

This object supports variable-size input signals, that is, the frame length (number of rows) of the signal can change even when the object is locked. However, the number of channels (columns) must remain constant.

Data Types: single | double
Complex Number Support: Yes

Output Arguments

y — Filtered output

vector | matrix

Output of the complex bandpass decimator, returned as a vector or a matrix. The output contains the subband of frequencies specified by the System object properties.

When the input is of size P-by-Q, and P is not a multiple of the decimation factor D, the output signal has an upper bound size of ceil(P/D)-by-Q, where D is the decimation factor. If P is a multiple of the decimation factor, then the output is of size (P/D)-by-Q. The number of channels (columns) does not change.

The data type of the output is same as the data type of the input. The output signal is always complex.

Data Types: single | double
Complex Number Support: Yes

Object Functions

To use an object function, specify the System object as the first input argument. For example, to release system resources of a System object named obj, use this syntax:

release(obj)

Specific to dsp.ComplexBandpassDecimator

cost	Implementation cost of the complex bandpass decimator
freqz	Frequency response of the multirate multistage filter
info	Information about filter System object
visualize	Visualize filter stages
filterAnalyzer	Analyze filters with Filter Analyzer app

Common to All System Objects

step	Run System object algorithm
release	Release resources and allow changes to System object property values and input characteristics
reset	Reset internal states of System object

Examples

Compute Cost of Complex Bandpass Decimator

Compute the implementation cost of a complex bandpass decimator using the cost function.

Create a dsp.ComplexBandpassDecimator object. Set the DecimationFactor to 12, center frequency to 0.2, and the NormalizedFrequency to true.

cbp = dsp.ComplexBandpassDecimator(12,0.2,NormalizedFrequency=true)

cbp = 
  dsp.ComplexBandpassDecimator with properties:

                CenterFrequency: 0.2000
                  Specification: 'Decimation factor'
               DecimationFactor: 12
            StopbandAttenuation: 80
                TransitionWidth: 0.0045
    MinimizeComplexCoefficients: true
            NormalizedFrequency: true

Compute the implementation cost of cbp using the cost function.

c = cost(cbp)

c = struct with fields:
                      NumCoefficients: 201
                            NumStates: 379
    RealMultiplicationsPerInputSample: 44.3333
          RealAdditionsPerInputSample: 43.8333

Compute Frequency Response of Complex Bandpass Decimator

Compute the complex frequency response of a complex bandpass decimator using the freqz function.

Create a dsp.ComplexBandpassDecimator object. Set the DecimationFactor to 12, the CenterFrequency to 5000 Hz, and the SampleRate to 44100 Hz. Compute and display the frequency response.

cbp = dsp.ComplexBandpassDecimator(12,5000,44100);
[h,f] = freqz(cbp);
plot(f,20*log10(abs(h)))
grid on
xlabel('Frequency (Hz)')
ylabel('h (dB)')

Filter Signal Through Complex Bandpass Decimator

Filter an input signal through a complex bandpass decimator and visualize the filtered spectrum in a spectrum analyzer.

Initialization

Create a dsp.ComplexBandpassDecimator System object™ with center frequency set to 2000 Hz, bandwidth of interest set to 1000 Hz, and sample rate set to 48 kHz. The decimation factor is computed as the ratio of the sample rate to the bandwidth of interest. The input to the decimator is a sine wave with a frame length of 1200 samples with tones at 1625 Hz, 2000 Hz, and 2125 Hz. Create a spectrumAnalyzer scope to visualize the signal spectrum.

Fs = 48e3;
CF = 2000;
BW = 1000;
D  =  Fs/BW;
FrameLength = 1200;
bpdecim = dsp.ComplexBandpassDecimator(D,CF,Fs);

sa = spectrumAnalyzer(SampleRate=Fs/D,...
    Method='welch',...
    YLimits=[-120 40],...
    FrequencyOffset=CF);

tones = [1625 2000 2125];
sin = dsp.SineWave(SampleRate=Fs,Frequency=tones,...
    SamplesPerFrame=FrameLength);

Visualize Filter Stages

Using the visualize function, you can visualize the response of each individual filter stage.

visualize(bpdecim)

Display Filter info

The info function displays information about the bandpass decimator.

fprintf('%s',info(bpdecim))

Overall Decimation Factor       : 48
Bandwidth                       : 1000 Hz
Number of Filters               : 5
Real multiplications per Input Sample: 14.708333
Real additions per Input Sample      : 13.833333
Number of Coefficients               : 89
Filters:                         
   Filter 1:
   dsp.FIRDecimator     - Decimation Factor   : 2 
   Filter 2:
   dsp.FIRDecimator     - Decimation Factor   : 2 
   Filter 3:
   dsp.FIRDecimator     - Decimation Factor   : 2 
   Filter 4:
   dsp.FIRDecimator     - Decimation Factor   : 3 
   Filter 5:
   dsp.FIRDecimator     - Decimation Factor   : 2 

Stream In and Filter Signal

Construct a for-loop to run for 1000 iterations. In each iteration, stream in 1200 samples (one frame) of the noisy sine wave and apply the complex bandpass decimator on each frame of the input signal. Visualize the input and output spectrum in the spectrum analyzer, sa.

for index = 1:1000
    x = sum(sin(),2) +  1e-4*randn(FrameLength,1);
    z = bpdecim(x);
    sa(z);
end

The bandpass decimator with center frequency at 2000 Hz and a bandwidth of 1000 Hz passes the three sine wave tones at 1625 Hz, 2000 Hz, and 2125 Hz.

Change the center frequency of the decimator to 2400 Hz and filter the signal.

release(bpdecim);
bpdecim.CenterFrequency = 2400

bpdecim = 
  dsp.ComplexBandpassDecimator with properties:

                CenterFrequency: 2400
                  Specification: 'Decimation factor'
               DecimationFactor: 48
            StopbandAttenuation: 80
                TransitionWidth: 100
    MinimizeComplexCoefficients: true
            NormalizedFrequency: false
                     SampleRate: 48000

Configure the spectrum analyzer to show the bandwidth of interest, [-1900, 2900] Hz.

release(sa)
sa.FrequencyOffset = 2400;

Stream in the data and filter the signal.

for index = 1:1000
    x = sum(sin(),2) +  1e-4 * randn(FrameLength,1);
    z = bpdecim(x);
    sa(z);
end

The tones at 2000 Hz and 2125 Hz are passed through the decimator, while the tone at 1625 Hz is filtered out.

Algorithms

The complex bandpass decimator is designed by applying a complex frequency shift transformation on a lowpass prototype filter. The lowpass prototype in this case is a multirate, multistage finite impulse response (FIR) filter. The desired frequency shift applies only to the first stage. Subsequent stages scale the desired frequency shift by their respective cumulative decimation factors. For details, see Complex Bandpass Filter Design and Zoom FFT.

Extended Capabilities

C/C++ Code Generation

Generate C and C++ code using MATLAB® Coder™.

Usage notes and limitations:

• See System Objects in MATLAB Code Generation (MATLAB Coder).

• This object also supports SIMD code generation using Intel^® AVX2 code replacement library under these conditions:

  • Input signal is complex-valued.

  • Input signal has a data type of single or double.

  The SIMD technology significantly improves the performance of the generated code. For more information, see SIMD Code Generation. To generate SIMD code from this object, see Use Intel AVX2 Code Replacement Library to Generate SIMD Code from MATLAB Algorithms.

• In the code you generate from this object, the decimator output is a fixed-size signal if the input signal is fixed size and is a multiple of the decimation factor. If the input is not a multiple of the decimation factor or if the input is a variable-size signal, then the decimator output is of variable-size.

Version History

Introduced in R2018a

R2024b: visualizeFilterStages has been renamed to visualize

The visualizeFilterStages function has been renamed to visualize. Existing instances of this function continue to run. For new instances, use visualize.

R2024b: visualize launches MATLAB figure

The visualize function now launches a MATLAB^® figure to visualize the filter stages of the complex bandpass decimator.

R2024b: Support for filter analyzer

The dsp.ComplexBandpassDecimator object now supports the filterAnalyzer function and the Filter Analyzer app.

R2024b: Support for normalized frequencies

When you set the NormalizedFrequency property to true, you must specify the center frequency, bandwidth, and the transition width in normalized frequency units (0 to 1). For more information, see the NormalizedFrequency property description.

R2023a: Support for arbitrary input frame length

This object supports an input signal with an arbitrary frame length, so the input frame length does not have to be a multiple of the decimation factor.