EEEN40130 Advanced Signal Processing

Academic Year 2023/2024

Building on the core principles established in the precursor module, EEEN30050 (Signal Processing), this module delves further into the domain of statistical signal processing, with a particular emphasis on random signals. Whereas EEEN30050 focuses on deterministic signals and their transformations, this module introduces students to properties of random signals and some fundamental statistical signal processing techniques such as power spectrum estimation, linear adaptive filtering, and some basic operations on image signals, such as template matching.

The module structure is as follows:

1. Random Signal Characterization: The module begins with the fundamentals of random signal characterization. This foundation lays the groundwork for understanding the nature of random signals and the statistical tools used to analyze them.

2. Estimation Theory Introduction: Students will be introduced to the basics of estimation theory. This portion provides the necessary concepts for predicting and understanding the behavior of random signals.

3. Filtering of Random Signals and Power Spectral Density Estimation: The next phase dives into the practical aspects of processing random signals. Students will explore different techniques for filtering random signals and estimating power spectral density. This section also includes optimal linear prediction and linear adaptive filtering.

4. Applications of Linear Filters: Once a firm understanding of the techniques is established, the module extends to the theory and applications of linear filters in various fields. This real-world application allows students to see the practicality of the theories learned.

5. Image Processing: The final segment of the module focuses on image processing. Students will learn how to apply signal processing techniques and domain-specific knowledge to derive efficient approaches for processing this unique type of signals.

Teaching methods are designed to be interactive, combining in-class lectures and tutorials with computer-based demonstrations. Students are encouraged to bring their laptops to each class, enabling them to engage actively with the numerical experiments conducted. All module materials will be made available at the beginning of the trimester, providing a comprehensive set of resources. Although a reading list is provided, the lecture notes are designed to be self-sufficient. Python is the programming language for this module.

The module is a blend of theory and practice, with real-world applications examined in depth. As part of the course requirements, students will be asked to confirm theoretical results by writing their own Python code. This will involve three comprehensive signal processing assignments. The deadlines for these tasks will be communicated well in advance, with a clear grading scheme outlined in the assignment handouts. A dedicated tutorial will be scheduled for each assignment, providing interactive, hands-on instruction. Each student will receive detailed, individual feedback for each task. In addition to the assignments, the assessment includes an 2 hour end-of-trimester exam.

Show/hide contentOpenClose All

Curricular information is subject to change

Learning Outcomes:

Having passed this module the student will be able to:

- Expound on the theoretical foundations of a number of important signal processing operations.

- Define fundamental terms that characterise a random signal.

- Describe and analyse linear estimators.

- Specify and design algorithms for the processing of speech and images and implement those algorithms in a high level language.

- Design, implement and analyse adaptive digital filters for a variety of applications.

Indicative Module Content:

The following key topics are covered:

- Random signals: random variables, stationary random processes, auto-correlation, cross-correlation, power spectral density.

- Estimation theory: unbiased estimation, minimum variance unbiased estimation, best linear unbiased estimations (BLUE), least squares.

- Linear optimum filtering: Wiener filter, linear prediction.

- Linear adaptive filtering: steepest-descent algorithm, least-mean-square algorithm.

- Channel equalization

- Image processing: image representation, 2D-convolution, 2D-correlation, 2D discrete Fourier transform, template matching, image filtering (smoothing filters, sharpening filters)

Student Effort Hours: 
Student Effort Type Hours
Lectures

48

Specified Learning Activities

25

Autonomous Student Learning

50

Total

123

Approaches to Teaching and Learning:
-Lectures, tutorials
-Problem-based learning 
Requirements, Exclusions and Recommendations
Learning Requirements:

Signal Processing (EEEN30050) or equivalent. This course is mathematically challenging, so a strong background in university honours level mathematics in the areas of linear algebra, frequency analysis (transforms), linear time invariant systems is required.


Module Requisites and Incompatibles
Not applicable to this module.
 
Assessment Strategy  
Description Timing Open Book Exam Component Scale Must Pass Component % of Final Grade
Assignment: Channel Equalization Varies over the Trimester n/a Graded No

15

Assignment: Image Processing Varies over the Trimester n/a Graded No

20

Assignment: Spectral Power Density Estimation Varies over the Trimester n/a Graded No

15

Examination: End of trimester exam 2 hour End of Trimester Exam No Standard conversion grade scale 40% No

50


Carry forward of passed components
No
 
Resit In Terminal Exam
Spring Yes - 2 Hour
Please see Student Jargon Buster for more information about remediation types and timing. 
Feedback Strategy/Strategies

• Feedback individually to students, post-assessment
• Group/class feedback, post-assessment

How will my Feedback be Delivered?

Not yet recorded.

- Digital Signal Processing: Principles, Algorithms, and Applications by John Proakis and Dimitris Manolakis
- Fundamentals of Statistical Signal Processing: Estimation Theory by Steven M. Kay.
- Adaptive Filter Theory by Simon Haykin
- Spectral analysis of signals by Peter Stoica and Randolph Moses
Name Role
Assoc Professor Nam Tran Lecturer / Co-Lecturer
Timetabling information is displayed only for guidance purposes, relates to the current Academic Year only and is subject to change.
 
Autumn
     
Lecture Offering 1 Week(s) - Autumn: All Weeks Fri 13:00 - 14:50
Lecture Offering 1 Week(s) - Autumn: All Weeks Tues 11:00 - 12:50
Autumn