On 15 Jun, 15:26, "mmoctar" <mmoc...@gmail.com> wrote:
> Is MUSIC algorithm interesting to estimate (amplitude and frequency) of
> sinusoids corrupted by noise?

MUSIC was designed to be a frequency estimator,
so it can be used to estimate frequencies. If
you know what you are doing, you can use the
frequency estimates to estimate the corresponding
amplitudes; just be aware that there is lots of
superstition and mythology surrounding MUSIC and
some of the derived results.

Whether MUSIC is 'interesting' for these types
of tasks depend entirely on the context of the
analysis. MUSIC has certain shortfalls, and
requires certain conditions to be guaranteed
up front. You basically need to know these
shortfalls and constraints, and find out whether
they apply to your particular situation.

Rune

On Jun 15, 9:26&#4294967295;am, "mmoctar" <mmoc...@gmail.com> wrote:
> Is MUSIC algorithm interesting to estimate (amplitude and frequency) of
> sinusoids corrupted by noise?
> Others propositions are welcome.
> Thank you in advance...
> Moctar

Depends on the situation, but sure, MUSIC or ESPRIT or
variations of them are interesting for line spectrum estimation
or signals that are sums or exponentials.
Julius

mmoctar wrote:
> Is MUSIC algorithm interesting to estimate (amplitude and frequency) of
> sinusoids corrupted by noise?
> Others propositions are welcome.
> Thank you in advance...
> Moctar

Is MUSIC algorithm interesting to estimate (amplitude and frequency) of
sinusoids corrupted by noise?
Others propositions are welcome.
Thank you in advance...
Moctar