Monday, January 21, 2013

Amplitude spectrum of a signal using Fourier transform

I have developed a MATLAB function to calculate amplitude spectrum of a signal using Fourier transform. Example usage of the function and implementation is shown below. Another example MATLAB program using built in function fft is also demonstrated. I hope they will be helpful for those who want to find amplitude spectrum or power spectrum of a signal.

%Example MATLAB code to test function FX
Fs=100;
t=(0:1/Fs:1-1/Fs)';
x=2*cos(2*pi*5*t)+sin(2*pi*10*t);
[f a]=FX(x,Fs);
plot(f,a);

function [f a]=FX(x,fs)
%Calculates amplitude spectrum of a signal x using Fourier transform
%[f a]=FX(x,fs)
%Input: x=signal, fs = sampling rate
%Output: f = frequency axis, a = amplitude spectrum
%File name: FX.m
%Author: Yan Naing Aye
%Website: http://cool-emerald.blogspot.sg/

dT = 1/fs; 
N=length(x);
dF=1/(N*dT);
NF=fs/2;%Nyquist Freq
%You can always limit freq range for faster performance
%NF=20;
t=(0:dT:(N-1)*dT)';
f=(0:dF:NF)';
a=f;%initialize a
a(1)=mean(x);
for i=2 : length(a)
    b=(2*mean(x.*cos(2*pi*dF*(i-1)*t)));
    c=(2*mean(x.*sin(2*pi*dF*(i-1)*t)));
    a(i)=sqrt(b^2+c^2);
end

%MATLAB program to calculates amplitude spectrum of a signal x using fft
%Author: Yan Naing Aye
%Website: http://cool-emerald.blogspot.sg/

Fs=100;
t=(0:1/Fs:1-1/Fs)';
x=2*cos(2*pi*5*t)+sin(2*pi*10*t);
L=length(x);

A=2*abs(fft(x)/L);
A=A(1:Fs/2+1);
f = Fs/2*linspace(0,1,Fs/2+1);
plot(f,A);

No comments:

Post a Comment

Comments are moderated and don't be surprised if your comment does not appear promptly.