Scribd
Upload
18 views

Exercise 4 MATLAB

This document contains instructions for a MATLAB exercise involving signal processing and parameter estimation. It provides three tasks: 1) implement MLE estimation for parameters in a cosine signal model and check variance with Monte Carlo simulation, 2) implement LSE estimation for frequency and phase in the same model using a phase locked loop, and 3) find LSE estimation for parameters in a piecewise signal model and compare to a constrained approach.

Uploaded by

Sumra Malik
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
18 views

Exercise 4 MATLAB

This document contains instructions for a MATLAB exercise involving signal processing and parameter estimation. It provides three tasks: 1) implement MLE estimation for parameters in a cosine signal model and check variance with Monte Carlo simulation, 2) implement LSE estimation for frequency and phase in the same model using a phase locked loop, and 3) find LSE estimation for parameters in a piecewise signal model and compare to a constrained approach.

Uploaded by

Sumra Malik
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

521348S Statistical Signal Processing I

Exercise 4

1. For the signal model

M M
x[n] = Ai cos(2πfi Ts n) + Bi sin(2πfi Ts n) + w[n]
X X

i=1 i=1

where n = 0, 1, ..., N − 1 with white Gaussian noise w[n] of variance σ 2 . Let sampling
period Ts = 1, frequencies fi are known with fi = i/N , and 2M < N . The amplitudes
θ = [A1 , A2 , ..., AM , B1 , B2 , ..., BM ] are to be estimated. Hint: The column of H are
orthogonal for the given frequencies, then H T H = N2 I.

(a) Define the entries of H and derive LSE of θ.


(b) Give the expressions of LSE for Âi and B̂i , and find minimum LSE.

2. Let us consider the signal model

0⩽n⩽K −1
(
θ,
s[n] =
−3θ, K ⩽ n ⩽ L − 1

(a) Find the LSE of θ and the corresponding minimum squared error.
(b) Assume further that data x[n] = s[n] + w[n] for n = 0, 1, 2, ...L − 1 are observed.
If now w[n] is zero-mean WGN with variance σ 2 , find the PDF of the LSE.

3. If the signal model is


s[n] = A + B(−1)n
where n = 0, 1, ..., N − 1 and N is even.

(a) Find the LSE of θ = [A B]T .


(b) Now, assume that A = B and repeat the problem using a constrained LS ap-
proach. Compare the results.

1
Matlab Exercise 3
Instructions*Please read carefully*
Return your solution to Moodle before 11.10.2022 10:30. Late returning will
lead to deduction of points. It is strongly recommended not to leave returning
to last minute but to do this well before the deadline.
Upload your solutions to Moodle as ZIPPED file containing all files needed
to run your codes. Your codes must be ready to run without any additional user
configuration.
Use filename M AT LAB_Group(#) . In addition to the well-commented program, make
sure the report includes the following:

• The full name and the student number for each group member.

• Clear answers to the problems posted (including for example figures obtained by run-
ning the simulation program).

Tasks
1. The data

x[n] = A cos(2πfo n + ϕ) + w[n]

for n = 0, 1, 2, ...N − 1 where A > 0 and 0 < fo < 1/2. Multiple parameters A, fo and
ϕ are unknown. Implement MLE estimator for θ = [A fo ϕ]T by minimizing

N −1
J(A, fo , ϕ) = (x[n] − A cos(2πfo n + ϕ))2
X

i=0

Implement Monte Carlo Simulation for MLE estimator to check their variance satisfies
the CRLB.
Hint: See Example 3.14 and Example 7.16 for variance and estimator equations.
Use these parameter values:
N = 5;
A = 1;
σ 2 = 0.001;
fo = 1/4;
ϕ = π/3;
M C = 100000; % number of Monte Carlo loops

2
2. For the same data

x[n] = A cos(2πfo n + ϕ) + w[n]

with n = −M, .., 0, .., M .


Implement LSE estimator for frequency fo and phase ϕ as a phase locked-loop with 10
times iteration.
Hint: See Example 8.14 for estimator equations.
Use these parameter values:
M = 2;
A = 1;
σ 2 = 0.001;
fo = 1/4;
ϕ = π/3;
Iteration = 10;

You might also like