Scribd
Upload
20 views

Practice Problem Set 2

1. The document provides 11 practice problems related to statistical signal processing. The problems cover topics like estimating parameters from independent and identically distributed samples, calculating the Cramer-Rao lower bound, finding maximum likelihood estimators, and determining if estimators are unbiased or efficient. 2. Some of the specific problems addressed include finding an unbiased estimator for data uniformly distributed between 0 and θ, determining if averaging improves estimation of heart rate measurements, and calculating the CRLB and determining if the minimum variance unbiased estimator exists for models with normal noise. 3. Estimation techniques covered include maximum likelihood, best linear unbiased, and minimum variance approaches. Distributions modeled include normal, uniform, and exponential. Parameters estimated

Uploaded by

abhay kumar
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
20 views

Practice Problem Set 2

1. The document provides 11 practice problems related to statistical signal processing. The problems cover topics like estimating parameters from independent and identically distributed samples, calculating the Cramer-Rao lower bound, finding maximum likelihood estimators, and determining if estimators are unbiased or efficient. 2. Some of the specific problems addressed include finding an unbiased estimator for data uniformly distributed between 0 and θ, determining if averaging improves estimation of heart rate measurements, and calculating the CRLB and determining if the minimum variance unbiased estimator exists for models with normal noise. 3. Estimation techniques covered include maximum likelihood, best linear unbiased, and minimum variance approaches. Distributions modeled include normal, uniform, and exponential. Parameters estimated

Uploaded by

abhay kumar
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

EE513L: Statistical Signal Processing (Jan-May 2023)

Practice Problem Set 2

1. Consider the data {x[0], x[1], . . . , x[N − 1]} where each sample is distributed as uniform random
variable between 0 and θ and samples are IID. Can you find an unbiased estimator for θ?
2. The heart rate h of a patient is automatically recorded by a computer every 100 msec which are
uncorrelated with each other. In 1 sec the measurements are averaged to obtain ĥ. If E(ĥi ) = αh for
some constant α and V ar(ĥi ) = 1 for each i, determine whether averaging improves the estimator
if α = 1 and α = 0.5.
3. Consider x[n] = θ + w[n], n = 0, 1 . . . , N − 1 be observations with w[n] ∼ N (0, β). Assume w[n]s
are IID. We want to estimate both θ, β. Is the following estimator unbiased?

N −1
1 X
θ̂ = x[n]
N n=0
N −1
1 X
β̂ = (x[n] − θ̂)2 .
N − 1 n=0

Find the MLE for θ, β.


4. The samples x[n] = rn + w[n], n = 0, 1, . . . , N − 1 are observed. Here w[n] ∼ N (0, σ 2 ) are IID. Find
the CRLB if r and σ 2 both are to be estimated. Does MVU exist? If yes, then find its variance.

5. The samples x[n] = θ + w[n], n = 0, 1, . . . , N − 1 are observed. Here w


⃗ ∼ N (0, C). Find the CRLB
for θ. Does MVU exist? If yes, then find its variance. Simplify the problem when you observe only
two samples.
6. Consider the generalization of line fitting example discussed in the class. Here we want to fit a
second order polynomial on the observed data with model: x[n] = A + Bn + Cn2 + w[n], n =
0, . . . , N − 1. Here w[n] ∼ N (0, σ 2 ). Find the CRLB for estimating A, B, C.
7. The samples x[n] = Arn + w[n], n = 0, 1, . . . , N − 1 are observed. Here w[n] ∼ N (0, σ 2 ) are IID and
r is known. Find BLUE for A and the minimum variance. Does the minimum variance approach
zero as n → ∞?

8. The samples x[n], n = 0, 1, . . . , N − 1 are observed. The samples are IID with distribution:
1
p(x[n]; θ) = exp(−|x[n] − θ|).
2
Find the BLUE of θ. What can you say about the MVU estimator in this case?
9. Find the MLE of θ when x[n] = A cos(2πf0 n + θ) + w[n], n = 0, 1, . . . , N − 1. Here w[n] ∼ N (0, σ 2 )
are IID.
10. We observe N idependent samples of a white Gaussian noise with variance σ 2 . We want to estimate
its power in dB which is given by P = 10 log10 σ 2 . Find the MLE of P.

11. For N IID observations from an uniform distribution PDF between 0 and θ, find MLE to estimate
θ.
12. For N IID observations from an exponential distribution with paramter λ find MLE to estimate λ.

You might also like