dsphdl.Channelizer

Polyphase filter bank and fast Fourier transform

Description

The dsphdl.Channelizer System object™ separates a broadband input signal into multiple narrowband output signals. It provides hardware speed and area optimization for streaming data applications. The object accepts scalar or vector input of real or complex data, provides hardware-friendly control signals, and has optional output frame control signals. You can achieve gigasamples-per-second (GSPS) throughput by using vector input. The object implements a polyphase filter, with one subfilter per input vector element. The hardware implementation interleaves the subfilters, which results in sharing each filter multiplier (FFT Length / Input Size) times. The object implements the same pipelined Radix 2^2 FFT algorithm as the dsphdl.FFT System object.

To channelize input data:

1. Create the dsphdl.Channelizer 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?

Note

You can also generate HDL code for this hardware-optimized algorithm, without creating a MATLAB^® script, by using the DSP HDL IP Designer app. The app provides the same interface and configuration options as the System object.

Creation

Syntax

channelizer = dsphdl.Channelizer

channelizer = dsphdl.Channelizer(Name=Value)

Description

channelizer = dsphdl.Channelizer returns a System object, channelizer, that implements a raised-cosine filter and an 8-point FFT.

channelizer = dsphdl.Channelizer(Name=Value) sets properties using one or more name-value arguments.

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.

NumFrequencyBands — FFT length

8 (default) | integer power of two

FFT length, specified as an integer power of two. For HDL code generation, the FFT length must be between 2² and 2¹⁶, inclusive.

FilterCoefficients — Polyphase filter coefficients

[-0.032 0.121 0.318 0.482 0.546 0.482 0.318 0.121 -0.032] (default) | vector of real or complex numeric values

Polyphase filter coefficients, specified as a vector of numeric values. If the number of coefficients is not a multiple of NumFrequencyBands, the object pads this vector with zeros. The default filter specification is a raised-cosine FIR filter, rcosdesign(0.25,2,4,'sqrt'). You can specify a vector of coefficients or a call to a filter design function that returns the coefficient values. By default, the object casts the coefficients to the same data type as the input.

FilterStructure — Filter design properties or coefficients

'Direct form transposed' (default) | 'Direct form systolic'

Specify the HDL filter architecture as one of these structures:

• Direct form transposed — This architecture is a fully parallel implementation that is suitable for FPGA and ASIC applications. For architecture and performance details, see Fully Parallel Transposed Architecture.

• Direct form systolic — This architecture provides a fully parallel filter implementation that makes efficient use of Intel^® and Xilinx^® DSP blocks. For architecture and performance details, see Fully Parallel Systolic Architecture.

ComplexMultiplication — HDL implementation of complex multipliers

'Use 4 multipliers and 2 adders' (default) | 'Use 3 multipliers and 5 adders'

HDL implementation of complex multipliers, specified as either 'Use 4 multipliers and 2 adders' or 'Use 3 multipliers and 5 adders'. Depending on your synthesis tool and target device, one option may be faster or smaller.

OutputSize — Size of output data

'Same as number of frequency bands' (default) | 'Same as input size'

Size of output data, specified as:

• 'Same as number of frequency bands' — Output data is a 1-by-M vector, where M is the FFT length.

• 'Same as input size' — Output data is an M-by-1 vector, where M is the input vector size.

The output order is bit natural for both output sizes.

Normalize — FFT scaling

true (default) | false

FFT output scaling, specified as either:

• true — The FFT implements an overall 1/N scale factor by scaling the result of each pipeline stage by 2. This adjustment keeps the output of the FFT in the same amplitude range as its input.

• false — The FFT avoids overflow by increasing the word length by one bit at each stage.

RoundingMethod — Rounding mode used for internal fixed-point calculations

'Floor' (default) | 'Ceiling' | 'Convergent' | 'Nearest' | 'Round' | 'Zero'

Rounding mode used for fixed-point operations. The object uses fixed-point arithmetic for internal calculations when the input is any integer or fixed-point data type. This option does not apply when the input is single or double. Each FFT stage rounds after the twiddle factor multiplication but before the butterflies. Rounding can also occur when casting the coefficients and the output of the polyphase filter to the data types you specify.

OverflowAction — Overflow handling for internal fixed-point calculations

'Wrap' (default) | 'Saturate'

Overflow Handling used for fixed-point operations. The object uses fixed-point arithmetic for internal calculations when the input is any integer or fixed-point data type. This option does not apply when the input is single or double. This option applies to casting the coefficients and the output of the polyphase filter to the data types you specify.

The FFT algorithm avoids overflow by either scaling the output of each stage (Normalize enabled), or by increasing the word length by 1 bit at each stage (Normalize disabled).

CoefficientsDataType — Data type of filter coefficients

'Same word length as input' (default) | numerictype object

The object casts the polyphase filter coefficients to this data type, using the rounding and overflow settings you specify. When you specify 'Same word length as input' (default), the object selects the binary point using fi() best-precision rules.

FilterOutputDataType — Data type of output of polyphase filter

'Same word length as input' (default) | 'Full precision' | numerictype object

Data type of the output of the polyphase filter, specified as 'Same word length as input', 'Full precision', or a numerictype object. The object casts the output of the polyphase filter (the input to the FFT) to this data type, using the rounding and overflow settings you specify. When you specify 'Full precision', the object selects a best-precision binary point by considering the values of your filter coefficients and the range of your input data type.

By default, the FFT logic does not change the data type. When you disable Normalize, the FFT algorithm avoids overflow by increasing the word length by 1 bit at each stage.

ResetInputPort — Enable reset argument

false (default) | true

Enable reset input argument to the object. When reset is 1 (true), the object stops the current calculation and clears internal states. When the reset is 0 (false) and the input valid is 1 (true), the object captures data for processing.

StartOutputPort — Enable start output argument

false (default) | true

Enable startOut output argument of the object. When enabled, the object returns an additional output signal that is 1 (true) on the first cycle of each valid output frame.

EndOutputPort — Enable end output argument

false (default) | true

Enable endOut output argument of the object. When enabled, the object returns an additional output signal that is 1 (true) on the first cycle of each valid output frame.

Usage

Syntax

[dataOut,validOut] = channelizer(dataIn,validIn)

[dataOut,validOut] = channelizer(dataIn,validIn,reset)

[dataOut,startOut,endOut,validOut] = channelizer(___)

Description

[dataOut,validOut] = channelizer(dataIn,validIn) filters and computes a fast Fourier transform, and returns the frequency channels, dataOut, detected in the input signal, dataIn, when validIn is 1 (true). The validIn and validOut arguments are logical scalars that indicate the validity of the input and output signals, respectively.

[dataOut,validOut] = channelizer(dataIn,validIn,reset) returns the frequency channels, dataOut, detected in the input signal, dataIn, when validIn is 1 (true) and reset is 0 (false). When reset is 1 (true), the object stops the current calculation and clears all internal state.

To use this syntax, set the ResetInputPort property to true. For example:

channelizer = dsphdl.Channelizer(...,ResetInputPort=true);
...
[dataOut,validOut] = channelizer(dataIn,validIn,reset)

[dataOut,startOut,endOut,validOut] = channelizer(___) returns the frequency channels, dataOut, computed from the input arguments of any of the previous syntaxes. startOut is 1 (true) for the first sample of a frame of output data. endOut is 1 (true) for the last sample of a frame of output data.

To use this syntax, set the StartOutputPort and EndOutputPort properties to true. For example:

channelizer = dsphdl.Channelizer(...,StartOutputPort=true,EndOutputPort=true);
...
[dataOut,startOut,endOut,validOut] = channelizer(dataIn,validIn)

Input Arguments

dataIn — Input data

scalar or column vector of real or complex values

Input data, specified as a scalar or column vector of real or complex values.

The vector size must be a power of 2 and in the range [2, 64], and is not greater than the number of channels (FFT length).

The software supports double and single data types for simulation, but not for HDL code generation.

The object does not accept uint64 data.

Data Types: fi | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | single | double
Complex Number Support: Yes

validIn — Indicates valid input data

scalar

Control signal that indicates if the input data is valid. When validIn is 1 (true), the object captures the values from the dataIn argument. When validIn is 0 (false), the object ignores the values from the dataIn argument.

Data Types: logical

reset — Clears internal states

logical scalar

Control signal that clears internal states. When reset is 1 (true), the object stops the current calculation and clears internal states. When the reset is 0 (false) and the input valid is 1 (true), the block captures data for processing.

For more reset considerations, see the Reset Signal section on the Hardware Control Signals page.

Dependencies

To enable this argument, set ResetInputPort to true.

Data Types: logical

Output Arguments

dataOut — Frequency channel output data

row vector

Frequency channel output data, returned as a row vector.

• If you set OutputSize to 'Same as number of frequency bands' (default), the output data is a 1-by-M vector, where M is the FFT length.

• If you set OutputSize to 'Same as input size', the output data is an M-by-1 vector, where M is the input vector size.

The output order is bit natural for either output size. The data type is a result of the FilterOutputDataType and the FFT bit growth necessary to avoid overflow.

validOut — Indicates valid output data

scalar

Control signal that indicates if the output data is valid. When validOut is 1 (true), the object returns valid data from the dataOut argument. When validOut is 0 (false), values from the dataOut argument are not valid.

Data Types: logical

startOut — First valid cycle of output data

logical scalar

First sample of output frame, returned as a logical scalar. The object sets startOut to 1 (true) during the first valid sample on dataOut.

Dependencies

To enable this argument, set StartOutputPort to true.

Data Types: logical

endOut — Last valid cycle of output data

logical scalar

Last sample of output frame, returned as a logical scalar. The object sets endOut to 1 (true) during the last valid sample on dataOut.

Dependencies

To enable this argument, set EndOutputPort to true.

Data Types: logical

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 dsphdl.Channelizer

getLatency

Latency of channelizer calculation

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

Create Channelizer for HDL Generation

Create a function that contains a channelizer object and supports HDL code generation.

Create the specifications and input signal. The signal has 8 frequency channels.

N = 8;
loopCount = 1024;
offsets = [-40 -30 -20 10 15 25 35 -15];
sinewave = dsp.SineWave(ComplexOutput=true, ...
    Frequency=offsets+(-375:125:500), ...
    SamplesPerFrame=loopCount);
sa = spectrumAnalyzer(ShowLegend=true, ...
     SampleRate=sinewave.SampleRate/N);

Write a function that creates and calls the channelizer System object™. You can generate HDL from this function.

function [yOut,validOut] = HDLChannelizer8(yIn,validIn)
%HDLChannelizer8 
% Process one sample of data using the dsphdl.Channelizer System object
% yIn is a fixed-point scalar or column vector. 
% validIn is a logical scalar value.
% You can generate HDL code from this function.

  persistent channelize8;
  coder.extrinsic('tf');
  coder.extrinsic('dsp.Channelizer');
  
  if isempty(channelize8)
    % Use filter coeffs from non-HDL channelizer, or supply your own.  
    channelizer = coder.const(dsp.Channelizer(NumFrequencyBands=8));
    coeff = coder.const(tf(channelizer));
    channelize8 = dsphdl.Channelizer(NumFrequencyBands=8,FilterCoefficients=coeff);
  end    
  [yOut,validOut] = channelize8(yIn,validIn);
end

% Copyright 2016-2023 The MathWorks, Inc.

Channelize the input data by calling the object for each data sample.

y = zeros(loopCount/N,N);
validOut = false(loopCount/N,1);
yValid = zeros(loopCount/(N*N),N);
for reps=1:10
   x = fi(sum(sinewave(),2),1,18);
   for loop=1:length(x)
      [y(loop,:),validOut(loop)]= HDLChannelizer8(x(loop),true);
   end
   yValid = y(validOut == 1,:);
   sa(yValid);
end

Explore Latency of Channelizer Object

The latency of the dsphdl.Channelizer object varies with the FFT length, filter structure, vector size, and input type. Use the getLatency function to find the latency of a particular configuration. The latency is measured as the number of cycles between the first valid input and the first valid output, assuming that the input is contiguous. The number of filter coefficients does not affect the latency. Setting the output size equal to the input size reduces the latency because the samples are not saved and reordered.

Create a dsphdl.Channelizer object with filter structure set to direct form transposed and request the latency.

channelize = dsphdl.Channelizer(NumFrequencyBands=512, ...
    FilterStructure='Direct form transposed');
L512 = getLatency(channelize)

L512 = 
1118

Request hypothetical latency information about a similar object with a different number of frequency bands (FFT length). The properties of the original object do not change.

L256 = getLatency(channelize,256)

L256 = 
592

N = channelize.NumFrequencyBands

N = 
512

Request hypothetical latency information of a similar object that accepts eight-sample vector input.

L256v8 = getLatency(channelize,256,8)

L256v8 = 
132

Enable scaling at each stage of the FFT. The latency does not change.

channelize.Normalize = true;
L512n = getLatency(channelize)

L512n = 
1118

Request the same output size and order as the input data. The latency decreases because the object does not need to store and reorder the data before output. The default input size is scalar.

channelize.OutputSize = 'Same as input size';
L512r = getLatency(channelize)

L512r = 
1084

Check the latency of a vector input implementation where the input and output are the same size. Specify the current value of the FFT length and a vector size of 8 samples. The latency decreases because the object computes results in parallel when the input is a vector.

L512rv8 = getLatency(channelize,channelize.NumFrequencyBands,8)

L512rv8 = 
218

Check the latency of a vector input implementation where the input type is complex. Specify the current value of the FFT length and a vector size of 16 samples.

L512rv16i = getLatency(channelize,channelize.NumFrequencyBands,16,true)

L512rv16i = 
152

Algorithms

This object implements the algorithm described on the Channelizer block reference page.

Note

The output of the dsphdl.Channelizer object does not match the output from the dsp.Channelizer object sample-for-sample. This mismatch is because the objects apply the input samples to the subfilters in different orders. The dsphdl.Channelizer object applies input X(0) to subfilter E_M-1(z), X(1) to subfilter E_M-2(z), ..., X(M-1) to subfilter E₀(z). The channels detected by both objects match, when analyzed over multiple frames.

Latency

The latency varies with the FFT length, vector size, and filter structure. Use the getLatency function to find the latency of a particular configuration. The latency is the number of cycles between the first valid input and the first valid output, assuming that the input is contiguous. The filter coefficients do not affect the latency. Setting the output size equal to the input size reduces the latency, because the samples are not saved and reordered.

Control Signals

This diagram shows validIn and validOut signals for contiguous input data with a vector size of 16, an FFT length of 512, and when you select the Direct form transposed filter architecture. In this example, the output vector size is specified same as the input vector size.

The diagram also shows the optional startOut and endOut signals that indicate frame boundaries. When enabled, startOut pulses for one cycle with the first validOut of the frame, and endOut pulses for one cycle with the last validOut of the frame.

If you apply continuous input frames (no gap in validIn between frames), the output will also be continuous, after the initial latency.

The validIn signal can be noncontiguous. Data accompanied by a validIn signal is stored until a frame is filled. Then the data in output is a contiguous frame of N/M cycles. This diagram shows noncontiguous input and contiguous output for an FFT length of 512 and a vector size of 16 samples.

Performance

These resource and performance data are the place-and-route results from the generated HDL targeted to a Xilinx Zynq^®- 7000 ZC706 evaluation board. The three examples in the tables use this common configuration.

• FFT length (default) — 8

• Filter length — 96 coefficients

• Filter structure — Direct form transposed

• 16-bit complex input data

• Coefficient data type — Same word length as input

• Filter output data type — Same word length as input

• Complex multiplication — Use 4 multipliers and 2 adders

• Output scaling — Enabled

• Output vector size — Same as input size

Performance of the synthesized HDL code varies with your target and synthesis options.

For scalar input, the design achieves a clock frequency of 506.84 MHz. The latency is 51 cycles. The subfilters share each multiplier eight (N) times. The design uses these resources.

Resource	Number Used
LUT	2898
FFS	3746
Xilinx LogiCORE® DSP48	28

For four-sample vector input, the design achieves a clock frequency of 452 MHz. The latency is 37 cycles. The subfilters share each multiplier twice (N/M). The design uses these resources.

Resource	Number Used
LUT	1991
FFS	8305
Xilinx LogiCORE DSP48	104

For eight-sample vector input, the design achieves a clock frequency of 360 MHz. The latency is 18 cycles. When the input size is the same as the FFT length, the subfilters do not share any multipliers. The design uses these resources.

Resource	Number Used
LUT	1683
FFS	2992
Xilinx LogiCORE DSP48	208

Extended Capabilities

C/C++ Code Generation

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

This System object supports C/C++ code generation for accelerating MATLAB simulations, and for DPI component generation.

HDL Code Generation

Generate VHDL, Verilog and SystemVerilog code for FPGA and ASIC designs using HDL Coder™.

The software supports double and single data types for simulation, but not for HDL code generation.

To generate HDL code from predefined System objects, see HDL Code Generation from Viterbi Decoder System Object (HDL Coder).

Version History

Introduced in R2017a

R2022a: Moved to DSP HDL Toolbox from DSP System Toolbox

Before R2022a, this System object was named dsp.HDLChannelizer and was part of the DSP System Toolbox™ product.

R2022a: FFT length of 4

You can now set the FFT length to 4 (2²). In previous releases the FFT length had to be a power of 2 from 8 (2³) to 2¹⁶.

R2022a: Direct form systolic filter structure support

This System object now supports fully parallel systolic architecture. To use this architecture, set the Filter structure property to 'Direct form systolic'.