addNoise

Add Gaussian noise to a signal for a given SNR.

Syntax

signal = addNoise(signal, snr)
signal = addNoise(signal, snr, mode)
[signal, snr] = addNoise(signal, snr)
[signal, snr] = addNoise(signal, snr, mode)

Description

addNoise adds Gaussian random noise to a one dimensional input signal given the desired signal snr (signal to noise ratio) in decibels. By default, the magnitude of the added noise is calculated based on the RMS level of the input signal. For impulsive signals, the optional input mode should be set to 'peak' so the magnitude of the added noise is calculated based on the peak level of the input signal. An example of adding noise to a sinusoidal signal is given below.

% specify signal properties
Fs = 50e6;          % [Hz]
dt = 1 / Fs;        % [s]
t = 0:dt:200 * dt;  % [s]

% create signal
signal = sin(0.5e6 * 2 * pi * t);

% add noise to give an snr of 20 dB and 10 dB
noisy_signal_a = addNoise(signal, 20);
noisy_signal_b = addNoise(signal, 10);

% plot
figure;
plot(t * 1e6, noisy_signal_a, 'r-', t * 1e6, noisy_signal_b, 'b-', t * 1e6, signal, 'k-');
xlabel('Time [\mus]');
ylabel('Signal [au]');
legend('20dB SNR', '10dB SNR', 'Original Signal');

Inputs

signal	input signal
snr	desired signal snr (signal to noise ratio) in decibels after adding noise

Optional Inputs

mode	'rms' (default) or 'peak'

Outputs

signal	signal with added noise
snr	snr of output signal

Examples

• 2D Time Reversal For A Circular Sensor
• Attenuation Compensation Using Time Reversal
• Attenuation Compensation Using Time Variant Filtering

See Also

randn