# comm.ChannelFilter

## Description

Use the `comm.ChannelFilter`

System object™ to filter a signal using multipath gains at specified path delays.

To filter a signal using multipath gains:

Create the

`comm.ChannelFilter`

object and set its properties.Call the object with arguments, as if it were a function.

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

## Creation

### Description

creates a
multipath channel filter System object to filter an input signal with path gains at the specified path
delays`chanFilt`

= comm.ChannelFilter

sets properties using one or more name-value pairs. For example,
`chanFilt`

= comm.ChannelFilter(`Name`

,`Value`

)`'SampleRate',1e6`

sets the sampling rate to `1`

MHz.
Enclose each property name in quotes.

## 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.

`SampleRate`

— Sample rate

`1`

Hz (default) | real positive scalar

Sample rate of the input signal, specified as a real, positive scalar.

**Data Types: **`double`

`PathDelays`

— Discrete path delays

`0`

(default) | real scalar | real vector

Delays of the discrete paths in seconds, specified as a real scalar or vector.

**Data Types: **`double`

`FilterDelaySource`

— Channel filter delay source

`'Auto'`

(default) | `'Custom'`

Channel filter delay source, specified as either `'Auto'`

or
`'Custom'`

.

Set

`FilterDelaySource`

to`'Auto'`

to specify the channel filter delay as the minimum possible value.Set

`FilterDelaySource`

to`'Custom'`

to specify the channel filter delay as a custom value. The custom value cannot be smaller than the minimum possible value.

**Data Types: **`char`

`FilterDelay`

— Channel filter delay

`7`

(default) | real non-negative integer scalar

Channel filter delay in samples, specified as a real, non-negative, integer scalar.

#### Dependencies

To enable this property, set the `FilterDelaySource`

property
to `'Custom'`

. The specified value must be no smaller than the
automatically determined channel filter delay when you set
`FilterDelaySource`

to `'Auto'`

.

**Data Types: **`double`

`NormalizeChannelOutputs`

— Normalize outputs by number of receive antennas

`1`

or `true`

(default) | `0`

or `false`

Normalize outputs by number of receive antennas, specified as a logical
`1`

(`true`

) or `0`

(`false`

).

**Data Types: **`logical`

## Usage

### Syntax

### Description

filters input signal `Y`

= chanFilt(`X`

,`G`

)`X`

, through a multipath channel with path gains
`G`

, at the path delay locations specified by the PathDelays
property.

### Input Arguments

`X`

— Input signal

matrix | `dlarray`

object

Input signal, specified as an
*N*_{S}-by-*N*_{T}
matrix, or a `dlarray`

(Deep Learning Toolbox)
object. *N*_{S} is the number of samples and
*N*_{T} is the number of transmit antennas. For
more information, see Array Support.

This object accepts variable-size inputs. After the object is locked, you can change the size of each input channel, but you cannot change the number of channels. For more information, see Variable-Size Signal Support with System Objects.

**Data Types: **`double`

| `single`

**Complex Number Support: **Yes

`G`

— Path gain

array

Path gain, specified as an array. For more information, see Array Support.

The input `G`

must be a
*N*_{S}-by-*N*_{P}-by-*N*_{T}-by-*N*_{R}
or
1-by-*N*_{P}-by-*N*_{T}-by-*N*_{R}
array, where *N*_{R} is the number of receive
antennas and *N*_{P} is the number of paths as
determined by the length of the PathDelays
property.

**Data Types: **`double`

| `single`

**Complex Number Support: **Yes

### Output Arguments

`Y`

— Channel output

matrix

Channel output, returned as a
*N*_{S}-by-*N*_{R}
matrix. The output matches the data type of `X`

. For more
information, see Array Support.

## 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)

## Examples

### Explore Spatial Diversity of Channel in Distributed MIMO System

In a distributed MIMO system, explore spatial diversity by transmitting the same signal from two geographically separated transmitters and combining the received signals at one receiver. Use ray tracing to analyze the propagation paths and gains from each transmitter to receiver.

**Perform Ray Tracing**

Import buildings data for Chicago into `siteviewer`

from an OpenStreetMap® (osm) file. For more information about the osm file, see [1]. Place two transmitter sites and one receiver site in the city.

sv = siteviewer("Buildings","Chicago.osm"); rx = rxsite("Name","Receiver", ... "Latitude",41.878543,"Longitude",-87.630599, ... "AntennaHeight",1.5); show(rx) tx1 = txsite("Name","Transmitter #1", ... "Latitude",41.878996,"Longitude",-87.629361); show(tx1) tx2 = txsite("Name","Transmitter #2", ... "Latitude",41.880142,"Longitude",-87.630850); show(tx2)

Perform ray tracing from the transmitter sites to the receiver site. Then, plot the propagation paths. By default, the `raytrace`

function uses the shooting and bouncing rays (SBR) method and calculate paths with up to two reflections.

rays = raytrace([tx1 tx2],rx); plot([rays{:}])

Ray tracing finds several ray paths to the receiver from each transmitter. From the map we can visually see the first transmitter is closer to the receiver than the second transmitter. We can also see the first transmitter has more reflected paths to the receiver. Display the propagation delays associated with each transmitter.

pd1 = [rays{1}.PropagationDelay]

pd1 =1×1010^{-6}× 0.3830 0.3839 0.5476 0.6482 0.5485 0.5486 0.6572 0.6945 0.7140 0.8874

pd2 = [rays{2}.PropagationDelay]

pd2 =1×510^{-6}× 0.5967 0.5973 0.6059 0.6066 0.6255

Construct one channel filter for each transmitter site. Specify a sample rate of 30 MHz and use the minimum delay among the seven rays as the reference of time 0.

chanFilt1 = comm.ChannelFilter( ... "SampleRate",30e6, ... "PathDelays",pd1-min([pd1, pd2]))

chanFilt1 = comm.ChannelFilter with properties: SampleRate: 30000000 PathDelays: [0 8.7598e-10 1.6459e-07 2.6516e-07 1.6553e-07 1.6565e-07 2.7417e-07 3.1151e-07 3.3105e-07 5.0442e-07] FilterDelaySource: 'Auto' NormalizeChannelOutputs: true

chanFilt2 = comm.ChannelFilter( ... "SampleRate",30e6, ... "PathDelays",pd2-min([pd1, pd2]))

chanFilt2 = comm.ChannelFilter with properties: SampleRate: 30000000 PathDelays: [2.1372e-07 2.1434e-07 2.2294e-07 2.2357e-07 2.4247e-07] FilterDelaySource: 'Auto' NormalizeChannelOutputs: true

The individual channel filters for the two transmitters yield different filter delay values. Use the `info`

object function of `comm.ChannelFilter`

to show the filter delay of the two channel filters.

fd1 = chanFilt1.info.ChannelFilterDelay

fd1 = 7

fd2 = chanFilt2.info.ChannelFilterDelay

fd2 = 1

The two channel filters must have the same filter delay to combine the channel outputs at the receiver site. Customize the filter delay for each channel filter to use the larger value of the individually computed delay values.

set(chanFilt1,"FilterDelaySource","Custom", ... "FilterDelay",max(fd1,fd2)); set(chanFilt2,"FilterDelaySource","Custom", ... "FilterDelay",max(fd1,fd2));

**Apply Receive Signal Combining**

Set up system parameters, assigning only one isotropic antenna at each site.

Nt = 1; % Number of transmit elements Ns = 1000; % Samples per frame M = 64; % Modulation order

Retrieve path gains from the computed rays. Assume the sites are static and no Doppler shift is introduced.

pg1 = 10.^(-[rays{1}.PathLoss]/20) .* ... exp(1i*[rays{1}.PhaseShift]); pg2 = 10.^(-[rays{2}.PathLoss]/20) .* ... exp(1i*[rays{2}.PhaseShift]);

Generate a frame of random 64-QAM signals. Perform channel filtering for each transmitter site and receive signal combining. The combined 2x1 distributed MIMO channel has a filter delay of max(`fd1`

,`fd2`

).

x = qammod(randi([0, M-1],Ns,Nt),M); y = chanFilt1(x,pg1) + chanFilt2(x,pg2);

**Appendix**

[1] The osm file is downloaded from https://www.openstreetmap.org, which provides access to crowd-sourced map data all over the world. The data is licensed under the Open Data Commons Open Database License (ODbL), https://opendatacommons.org/licenses/odbl/.

### Perform Channel Filtering for an LTE 2x2 EVA Profile

Construct a channel filter object with the LTE Extended Vehicular A model (EVA) delay profile.

chanFilt = comm.ChannelFilter( ... 'SampleRate', 30.72e6, ... 'PathDelays', [0 30 150 310 370 710 1090 1730 2510]*1e-9);

Set up system parameters. There are two transmit and receive antennas.

[Nt, Nr] = deal(2); Ns = 30720; Np = length(chanFilt.PathDelays); M = 256;

Generate random 256-QAM signal and complex path gains.

x = qammod(randi([0, M-1], Ns, Nt), M); g = complex(rand(Ns, Np, Nt, Nr), rand(Ns, Np, Nt, Nr));

Filter the signal with path gains for the EVA delay profile.

y = chanFilt(x, g);

### Reciprocal Downlink and Uplink Transmissions in MIMO Channel

Using one MIMO channel System object™ and two identically configured channel filter System objects, switch a link-level simulation between 3-by-2 downlink and reciprocal 2-by-3 uplink signal transmissions.

Define system parameters.

modOrder = 256; % Modulation order Nant1 = 3; % Number of 'transmit' antennas Nant2 = 2; % Number of 'receive' antennas Rs = 1e6; % Sample rate pd = [0 1.5 2.3]*1e-6; % Path delays frmLen = 1e3; % Frame length

Create a MIMO channel System object™, configuring it for path gain generation by disabling channel filtering.

chan = comm.MIMOChannel( ... 'SampleRate',Rs, ... 'PathDelays',pd, ... 'AveragePathGains',[1.5 1.2 0.2], ... 'MaximumDopplerShift',300, ... 'SpatialCorrelationSpecification','none', ... 'NumTransmitAntennas',Nant1, ... 'NumReceiveAntennas',Nant2, ... 'ChannelFiltering',false, ... 'NumSamples',frmLen);

Create identical channel filter System objects for both transmission directions: one channel filter for the `Nant1`

-by-`Nant2`

downlink channel (3 transmit antennas to 2 receive antennas) and a reciprocal channel filter for the `Nant2`

-by-`Nant1`

uplink channel (2 transmit antennas to 3 receive antennas).

chanFiltDownlink = comm.ChannelFilter( ... 'SampleRate',Rs, ... 'PathDelays',pd); chanFiltUplink = clone(chanFiltDownlink);

**Downlink Transmission**

Generate random path gains for one frame of the downlink 3-by-2 channel. Pass randomly generated 256-QAM signals through the 3-by-2 downlink channel.

pgDownlink = chan(); x = qammod(randi([0 modOrder-1],frmLen,Nant1),modOrder); yDL = chanFiltDownlink(x,pgDownlink);

**Uplink Transmission**

Switch the link direction. Run the channel object to generate another frame of path gains, permuting its 3rd (Tx) and 4th (Rx) dimensions for the reciprocal uplink 2-by-3 channel. Pass randomly generated 256-QAM signals through the 2-by-3 reciprocal uplink channel.

pgUplink = permute(chan(),[1 2 4 3]); x = qammod(randi([0 modOrder-1],frmLen,Nant2),modOrder); yUL = chanFiltUplink(x,pgUplink);

**Downlink and Uplink Array Dimensions**

Show the sizes of the downlink and uplink path gain arrays returned by the MIMI channel object as an ${\mathit{N}}_{\mathrm{S}}$-by-${\mathit{N}}_{\mathrm{P}}$-by-${\mathit{N}}_{\mathrm{T}}$-by-${\mathit{N}}_{\mathrm{R}}$ array.

${\mathit{N}}_{\mathrm{S}}$ is the number of samples.

${\mathit{N}}_{\mathrm{P}}$ is the number of path delays.

${\mathit{N}}_{\mathrm{T}}$ is the number of transmit antennas.

`Nant1`

for downlink and`Nant2`

for uplink.${\mathit{N}}_{\mathrm{R}}$ is the number of receive antennas.

`Nant2`

for downlink and`Nant1`

for uplink.

size(pgDownlink)

`ans = `*1×4*
1000 3 3 2

size(pgUplink)

`ans = `*1×4*
1000 3 2 3

Show the size of the channel output matrices returned by the MIMI channel object as an ${\mathit{N}}_{\mathrm{S}}$-by-${\mathit{N}}_{\mathrm{R}}$ matrix. ${\mathit{N}}_{\mathrm{S}}$ is the number of samples. ${\mathit{N}}_{\mathrm{R}}$ is the number of receive antennas.

size(yDL)

`ans = `*1×2*
1000 2

size(yUL)

`ans = `*1×2*
1000 3

## More About

### Array Support

The `comm.ChannelFilter`

object supports an input signal
represented in an array, `dlarray`

(Deep Learning Toolbox), or
`gpuArray`

(Parallel Computing Toolbox).

If

`X`

is specified as a`dlarray`

,`Y`

is returned as a`dlarray`

object.If

`X`

is a`dlarray`

holding a`gpuArray`

, then`G`

must be a`gpuArray`

.If

`X`

is a`dlarray`

holding an array, then`G`

must be an array.The number of batch observations (

*N*_{B}) is an optional dimension that can be added to the input for all supported data types. Variable*N*_{B}is not supported. When the*N*_{B}dimension is included:`X`

— The input dimensions can be an array of up to three dimensions, specified as an*N*_{S}-by-*N*_{T}-by-*N*_{B}array.`G`

— If`X`

is a 3-D array,`G`

can be an*N*_{S}-by-*N*_{P}-by-*N*_{T}-by-*N*_{R}or*N*_{S}-by-*N*_{P}-by-*N*_{T}-by-*N*_{R}-by-*N*_{B}array.`Y`

— The output is returned as an*N*_{S}-by-*N*_{R}matrix or an*N*_{S}-by-*N*_{R}-by-*N*_{B}array.

For a list of Communications Toolbox™ features that support `dlarray`

objects, see AI for Wireless.

## Algorithms

### Channel Filter Model Characteristics

The channel filter implements a fractional delay (FD) finite impulse response (FIR) bandpass filter with a length of 16 coefficients for each candidate fractional delay at 0, 0.02, 0.04, …, 0.98.

Each discrete path is rounded to its nearest candidate fractional delay, so the delay error limit is 1% of the sample time. To achieve a group delay bandwidth exceeding 80% and a magnitude bandwidth exceeding 90%, the algorithm selects the optimal FIR coefficient values for each fractional delay, while satisfying the following criteria:

Group delay ripple ≤ 10%

Magnitude ripple ≤ 2 dB

Magnitude bandedge attenuation = 3 dB

The plots show bandwidths that satisfy the design criteria for group delay ripple, magnitude ripple, and magnitude bandedge attenuation.

For additional information, see the article *A Matlab-based Object-Oriented Approach to Multipath Fading Channel Simulation* at MATLAB^{®} Central.

## 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).

### GPU Arrays

Accelerate code by running on a graphics processing unit (GPU) using Parallel Computing Toolbox™.

This System object supports GPU array inputs. For more information, see Accelerate Simulation Using GPUs.

Usage notes and limitations:

## Version History

**Introduced in R2020b**

### R2024b: Add deep learning array support

The `comm.ChannelFilter`

object adds support for `dlarray`

(Deep Learning Toolbox) object
processing for deep learning applications.

### R2024a: Add GPU array support

The `comm.ChannelFilter`

System object adds support for `gpuArray`

(Parallel Computing Toolbox) object processing to run code on a
graphics processing unit (GPU).

## See Also

### Objects

## MATLAB 명령

다음 MATLAB 명령에 해당하는 링크를 클릭했습니다.

명령을 실행하려면 MATLAB 명령 창에 입력하십시오. 웹 브라우저는 MATLAB 명령을 지원하지 않습니다.

Select a Web Site

Choose a web site to get translated content where available and see local events and offers. Based on your location, we recommend that you select: .

You can also select a web site from the following list:

## How to Get Best Site Performance

Select the China site (in Chinese or English) for best site performance. Other MathWorks country sites are not optimized for visits from your location.

### Americas

- América Latina (Español)
- Canada (English)
- United States (English)

### Europe

- Belgium (English)
- Denmark (English)
- Deutschland (Deutsch)
- España (Español)
- Finland (English)
- France (Français)
- Ireland (English)
- Italia (Italiano)
- Luxembourg (English)

- Netherlands (English)
- Norway (English)
- Österreich (Deutsch)
- Portugal (English)
- Sweden (English)
- Switzerland
- United Kingdom (English)