Existence of a (non-Gaussian) process with a given covariance function

[3/33%]
Show
LaTeX

Let r=r(s,t)r = r(s, t) be a nonnegative definite function (s,t∈T(s, t \in T, where TT is some set). Does there exist a non-Gaussian process X={Xt,t∈T}X = \left\{ X_{t}, t \in T\right\} such that cov⁡(Xs,Xt)=r(s,t)\operatorname {cov}\left(X_{s}, X_{t}\right) = r(s, t) for all s,t∈Ts, t \in T ?

?

Which of the following functions can be the covariance functions to a stationary process and which can not?

?
(a)

r(τ)=sin⁡ττr(\tau ) = \dfrac {\sin \tau }{\tau },

(b)

r(τ)=sin⁡ττ2r(\tau ) = \dfrac {\sin \tau }{\tau^2},

(c)

r(τ)={1,τ=0,2,τ=1,0.5,τ=2,0,otherwise,r(\tau ) = \begin{cases} 1, & \tau =0, \\ 2, & \tau =1, \\ 0.5, & \tau =2, \\ 0, & \text{otherwise}, \end{cases}

(d)

r(τ)=cos⁡τr(\tau ) = \cos \tau,

(e)

r(τ)=e∣τ∣r(\tau ) = e^{\left|\tau \right|}.

Determine whether the following functions are covariance functions of some Gaussian processes (Xt,t∈R)\left(X_{t}, t \in \mathbb {R}\right):

?
(a)

r(s,t)=e−α∣s−t∣,α>0r(s, t)=e^{-\alpha \left|s-t\right|}, \alpha >0;

(b)

r(s,t)=2cos⁡(t−s)r(s, t)=2 \cos (t-s);

(c)

r(s,t)=(1−(t−s)2)I{∣s−t∣≤1}r(s, t)=\left(1-(t-s)^{2}\right) I\left\{ \left|s-t\right| \leq 1\right\};

(d)

r(s,t)=exp⁡{ia(s−t)−(t−s)2σ22}r(s, t)=\exp \left\{ i a(s-t)-\frac{(t-s)^{2} \sigma^{2}}{2}\right\}.

For those processes that do exist, determine whether they have spectral densities.

References

[БуШи05] Булинский А.В., Ширяев А.Н. Теория случайных процессов. ФИЗМАТЛИТ, 2005.
[LRS14] Lindgren G., Rootzén H., Sandsten M. Stationary Stochastic Processes for Scientists and Engineers. CRC Press (Taylor & Francis Group), 2014.
[ЖРШ16] Жуковский М.Е., Родионов И.В., Шабанов Д.А. Основы теории случайных процессов. МФТИ, 2016.