IR deconvolution from cross-correlation and auto-correlation

Hello,

I am after getting the impulse response of a process.

To do so, it is common to give the process a measured time dependent input, in(t), and measure the time dependent output, out(t).

it is well known that the output is the convolution of the input and the impulse response of the system, i.e. out(t)=h(t)*in(t).

To get h(t) it is common to perform: h(t) = ifft(fft(out(t))/in(t)) however in many cases filters need to be used to smooth the signal as it contains higher and lower order harmonics originating from the noise.

A colleague suggested the use of a Yule-Walker approach to solve the r_{y,x} = h*r_{x,x} where r_{y,x} is the cross-correlation of the output and input and r_{xx} is the autocorrelation of the input signal. can anyone help on how that can be performed in Matlab?

Please keep in mind, that I am after h(t) - the time dependent response as I will further use it to determine certain parameters of the system.

Regards, F.Z.