How to solve Harmonic and noise model of speech signal, more in description

What I need to do is:

-Perform the decomposition of voiced sounds of speech signal into harmonic components. -Re-synthesized signal by using the harmonic components. -Perform scrambling and reverse synthesis unvoiced sounds of the speech signal based on LPC. -Implement this synthesis for non-stationary signal, based on harmonic model the only modeling voiced speech segments. Detection of voicing do based on energy detection and ZCR.

Answer 1

[Terrance Nearey]

Dan Ellis' code available at this link might get you started on sinusoidal analysis/resynthisis: https://labrosa.ee.columbia.edu/matlab/sinemodel/