Main Content

comm.gpu.AWGNChannel

Add white Gaussian noise to input signal with GPU

expand all in page

Description

The GPU AWGNChannel object adds white Gaussian noise to an input signal using a graphics processing unit (GPU).

Note

To use this object, you must install a Parallel Computing Toolbox™ license and have access to an appropriate GPU. For more about GPUs, see GPU Computing (Parallel Computing Toolbox).

To add white Gaussian noise to an input signal:

  1. Define and set up your additive white Gaussian noise channel object. See Construction.

  2. Call step to add white Gaussian noise to the input signal according to the properties of comm.gpu.AWGNChannel. The behavior of step is specific to each object in the toolbox.

A GPU-based System object™ accepts typical MATLAB® arrays or objects created using the gpuArray class. A GPU-based System object supports input signals with double- or single-precision data types. The output signal inherits its data type from the input signal.

  • If the input signal is a MATLAB array, the System object handles data transfer between the CPU and the GPU. The output signal is a MATLAB array.

  • If the input signal is a gpuArray, the data remains on the GPU. The output signal is a gpuArray. When the object is given a gpuArray, calculations take place entirely on the GPU, and no data transfer occurs. Passing gpuArray arguments provides increased performance by reducing simulation time. For more information, see Establish Arrays on a GPU (Parallel Computing Toolbox).

Note

Starting in R2016b, instead of using the step method to perform the operation defined by the System object, you can call the object with arguments, as if it were a function. For example, y = step(obj,x) and y = obj(x) perform equivalent operations.

Construction

H = comm.gpu.AWGNChannel creates a GPU-based additive white Gaussian noise (AWGN) channel System object, H. This object adds white Gaussian noise to a real or complex input signal.

H = comm.gpu.AWGNChannel(Name,Value) creates a GPU-based AWGN channel object, H, with the specified property name set to the specified value. You can specify additional name-value pair arguments in any order as (Name1,Value1,...,NameN,ValueN).

A GPU-based System object accepts typical MATLAB arrays or objects created using the gpuArray class. A GPU-based System object supports input signals with double- or single-precision data types. The output signal inherits its data type from the input signal.

  • If the input signal is a MATLAB array, the System object handles data transfer between the CPU and the GPU. The output signal is a MATLAB array.

  • If the input signal is a gpuArray, the data remains on the GPU. The output signal is a gpuArray. When the object is given a gpuArray, calculations take place entirely on the GPU, and no data transfer occurs. Passing gpuArray arguments provides increased performance by reducing simulation time. For more information, see Establish Arrays on a GPU (Parallel Computing Toolbox).

Properties

NoiseMethod

Method to specify noise level

Select the method to specify the noise level as one of Signal to noise ratio (Eb/No) | Signal to noise ratio (Es/No) | Signal to noise ratio (SNR) | Variance. The default is Signal to noise ratio (Eb/No).

EbNo

Energy per bit to noise power spectral density ratio (Eb/No)

Specify the Eb/No ratio in decibels. Set this property to a numeric, real scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Signal to noise ratio (Eb/No). The default is 10. This property is tunable.

EsNo

Energy per symbol to noise power spectral density ratio (Es/No)

Specify the Es/No ratio in decibels. Set this property to a numeric, real scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Signal to noise ratio (Es/No). The default is 10. This property is tunable.

SNR

Signal to noise ratio (SNR)

Specify the SNR value in decibels. Set this property to a numeric, real scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Signal to noise ratio (SNR). The default is 10. This property is tunable.

BitsPerSymbol

Number of bits in one symbol

Specify the number of bits in each input symbol. You can set this property to a numeric, positive, integer scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Signal to noise ratio (Eb/No). The default is 1 bit.

SignalPower

Input signal power in Watts

Specify the mean square power of the input signal in Watts. Set this property to a numeric, positive, real scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Signal to noise ratio (Eb/No), Signal to noise ratio (Es/No) or Signal to noise ratio (SNR). The default is 1 Watt. The object assumes a nominal impedance of 1 Ohm. This property is tunable.

SamplesPerSymbol

Number of samples per symbol

Specify the number of samples per symbol. Set this property to a numeric, positive, integer scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Signal to noise ratio (Eb/No) or Signal to noise ratio (Es/No). The default is 1 sample.

VarianceSource

Source of noise variance

Specify the source of the noise variance as one of Property | Input port. The default is Property. Set VarianceSource to Input port to specify the noise variance value via an input to the step method. Set VarianceSource to Property to specify the noise variance value using the Variance property. This property applies when you set the NoiseMethod property to Variance.

Variance

Noise variance

Specify the variance of the white Gaussian noise. You can set this property to a numeric, positive, real scalar or row vector with a length equal to the number of channels. This property applies when you set the NoiseMethod property to Variance and the VarianceSource property to Property. The default is 1. This property is tunable.

RandomStream

Source of random number stream

Specify the source of random number stream. The only valid setting for this property is Global stream. The object generates the normally distributed random numbers from the current global random number stream.

Seed

Initial seed of mt19937ar random number stream

The GPU version of the AWGN Channel System object does not use this property.

Methods

stepAdd white Gaussian noise to input signal
Common to All System Objects
release

Allow System object property value changes

Algorithm

This object uses the same algorithm as the comm.AWGNChannel System object. See the Algorithms section of the comm.AWGNChannel help page for more details. The object properties correspond to the related block parameters, except that:

  • This object uses parallel.gpu.RandStream to provide an interface for controlling the properties of one or more random number streams that the GPU uses. Usage is the same as RandStream with the following restrictions:

    • Only the combRecursive (MRG32K3A) generator is supported.

    • Only the Inversion normal transform is supported.

    • Setting the substream property is not allowed.

Enter help parallel.gpu.RandStream at the MATLAB command line for more information.

Examples

collapse all

Open Script

Specify the modulation order and generate PSK-modulated random data.

M = 8;
modData = pskmod(randi([0 M-1],1000,1),M,pi/M);

Create an AWGN channel object that uses a GPU. Pass the modulated data through the channel.

gpuChannel = comm.gpu.AWGNChannel('EbNo',15,'BitsPerSymbol', ...
                log2(M));
channelOutput = gpuChannel(modData);

Visualize the noiseless and noisy data in scatter plots.

scatterplot(modData)
scatterplot(channelOutput)

Extended Capabilities

Version History

Introduced in R2012a

See Also