The frequency domain is another way to obtain the weak (wide sense) characterization of a stationary stochastic process.
Consider a stationary stochastic process y(t):
E[y(t)]=my∀t
γy(τ)=E[(y(t)−my)(y(t−τ)−my)]=my
Then we define the spectrum as the discrete Fourier transform of the covariance function γy(τ) of the stationary stochastic process y(t):
Γ(ω):=τ=−∞∑+∞γy(τ)e−jωτ
Example: spectrum of the white noise
Given the white noise e(t)∼EN(μ,λ2) which has the covariance function:
γe(τ)={λ20if τ=0if τ=0
The spectrum is:
Γ(ω)=τ=−∞∑+∞γe(τ)e−jωτ=γe(τ)e−jω0=λ2
Example: spectrum of MA(1) process
Consider the process y(t)=e(t)+ce(t−1), his covariance function is:
γy(τ)=⎩⎨⎧(12+c2)λ2(1×c)λ20if τ=0if τ=±1if τ=0
The spectrum of the process is then:
Γ(ω)=τ=−∞∑+∞γy(τ)e−jωτ=(1+c2)λ2e−jω0+cλ2e−jω1+cλ2e−jω(−1)=λ2[1+c2+c(e−jω+ejω)]
Since we know that: ejω=cos(ω)+jsin(ω) and then e−jω=cos(ω)−jsin(ω)
We can rewrite the spectrum as: Γ(ω)=λ2(1+c2+2cos(ω))
Properties of the spectrum
The properties of the spectrum are inherited from the properties of γy(τ).
Γ(ω)∈R∀ω the spectrum is real valued. This happens, intuitivelly, because since γy(τ) = γy(−τ) then the immaginary parts cancel out.
Γ(ω)≥0∀ω the spectrum is positive.
Γ(ω)=Γ(−ω)∀ω the spectrum is an even function.
Γ(ω)=Γ(ω+2kπ)∀ω,k=0,±1,... the spectrum is periodic with period 2π.
Spectrum of output of digital filters
Given the well defined stationary stochastic process y(t)=W(z)v(t)=B(z)A(z)v(t) we can compute the covarianca function and then the spectrum.
There exists an easier way... an important theorem states that:
Γy(ω)=∣W(ejw)∣2Γv(ω)
Which means that the spectrum of the output is the spectrum of the input multiplied by the transfer function evaluated for z=ejw absolute value square.
Example: spectrum of MA(1) using digital filter th.
Given the MA(1) process y(t)=e(t)+ce(t−1)=(1+cz−1)e(t) with e(t)∼EN(μ,λ2) the spectrum of the process is:
Γy(ω)=∣W(ejw)∣2Γe(ω)=∣1+ce−jw∣2λ2=λ2(1+c2+2cos(ω))