Hi Siddharth,
As per my understanding, you want to remove the transient part in FFT.
To remove the transient part of FFT, you must apply a window function to the input signal before computing the FFT. Consider the code shown below:
N=128;
window_start=1;
window_stop=20;
x=rand(N,1);
x(window_start:window_stop)=x(window_start:window_stop).*hann(window_stop-window_start+1);
X=fft(x);
plot(abs(X));
where, "N" is the FFT size and "x" is a 128x1 matrix containing random values ranging in the range (0,1). To remove transient in a particular interval, "window_start" & "window_stop" are used.
The window function "hann" is applied to the values of "x" in the interval (window_start, window_stop), to remove the transient in that interval. Then the FFT is computed with the help of "fft" function.
For more information on the above-mentioned functions, you can refer to the following documentation:
For "rand" function:
For "hann" function:
For "fft" function:
