1. The expectation operator is linear, where c is a constant. If X and Y are jointly normal and zero mean, then their cross-correlation is determined by solving equations involving their variances and covariance. If X and Y are circularly symmetric and independent, then they are Gaussian with equal variance.
2. To determine the auto-correlation function of a random walk process X(t) with time interval T and step size S, consider the expected value of the product of X(t) at two times. If a stochastic process X(t) has a mean-square derivative X'(t), then its auto-correlation function satisfies certain properties involving its derivative.
3. For jointly Gaussian, stationary
1. The expectation operator is linear, where c is a constant. If X and Y are jointly normal and zero mean, then their cross-correlation is determined by solving equations involving their variances and covariance. If X and Y are circularly symmetric and independent, then they are Gaussian with equal variance.
2. To determine the auto-correlation function of a random walk process X(t) with time interval T and step size S, consider the expected value of the product of X(t) at two times. If a stochastic process X(t) has a mean-square derivative X'(t), then its auto-correlation function satisfies certain properties involving its derivative.
3. For jointly Gaussian, stationary
1. The expectation operator is linear, where c is a constant. If X and Y are jointly normal and zero mean, then their cross-correlation is determined by solving equations involving their variances and covariance. If X and Y are circularly symmetric and independent, then they are Gaussian with equal variance.
2. To determine the auto-correlation function of a random walk process X(t) with time interval T and step size S, consider the expected value of the product of X(t) at two times. If a stochastic process X(t) has a mean-square derivative X'(t), then its auto-correlation function satisfies certain properties involving its derivative.
3. For jointly Gaussian, stationary
1. The expectation operator is linear, where c is a constant. If X and Y are jointly normal and zero mean, then their cross-correlation is determined by solving equations involving their variances and covariance. If X and Y are circularly symmetric and independent, then they are Gaussian with equal variance.
2. To determine the auto-correlation function of a random walk process X(t) with time interval T and step size S, consider the expected value of the product of X(t) at two times. If a stochastic process X(t) has a mean-square derivative X'(t), then its auto-correlation function satisfies certain properties involving its derivative.
3. For jointly Gaussian, stationary
Download as DOCX, PDF, TXT or read online from Scribd
Download as docx, pdf, or txt
You are on page 1/ 3
1.
Give the proof for the following statements:
(1) The expectation operator is linear, i.e., 𝐄𝐌𝐁𝐄𝐃 where c is a constant (2) (3) if
2. Use the Price's Theorem to prove the following proposition:
If are jointly normal and zero mean, then
and
where a is determined by solving
in which and . Also
= bivariate normal
3. Give the proof for the Theorem:
If X and Y are circularly symmetric and independent, then X and Y are Gaussian with mean and equal variance
4. Consider a random walk process with time interval T and step size S and X(0)=0. Let
Question: how to determine the auto-correlation function of X(t)?
X (t ) ' X ( t ) , prove that 5. A SP has a m.s. derivative
∂2 R t ,t (1) X X 1 2 ∂ t ∂ t XX ( 1 2 ) R ' ' ( t ,t )=
1 2
d2 ′ R X ' X ' ( τ )=− R XX ( τ ) (2) if X (t ) is stationary, then dτ 2
6. Consider X ( t ) and to be jointly Gaussian, stationary, and zero mean.
Give the proof for the following statements:
Let (1) (2) (3) (4)
7. Prove the properties of auto-correlation function of random process :
¿ (1) R (−τ )=R ( τ )
(2) R ( 0 ) ≥0
(3) |Re [ R ( τ ) ]|≤R ( 0 )
n n is positive definite,i.e.∑ ∑ R ( τ m−τ l ) al a m≥0 ∀ a ,...,a ∀ τ ¿ Where R (τ ) l=1 m=1 1 m m
8. Consider two random process and ,
where is a constant and is uniformly distributed over any region.
(1) Find the means of and respectively
(2) Show that the cross-correlation function of and is given by (3) Suppose we sample and and at regular time intervals of duration T to form two sequences of random variables and respectively. Find the smallest value of T such that the two sequences are uncorrelated, i.e
9. The transition of wet (w) and dry (D) days in a city is modeled as a Markov chain
with transition probability matrix which is shown as follow , draw the
transition figure of that Markov chain.
If it is dry today, what is the probability that it will be dry 2 days from today?
10. Give the proof for the following statement:
If X(t) is Markoff and Gaussian and zero mean, then its autocorrelation must satisfy .
