---

ENGINEERING TRIPOS PART IIA – 2012/2013

Module 3F3 – Signal & Pattern Processing

Leader:

Prof. S. Godsill

Timing:	Lent
Prerequisites:	3F1
Structure:	16L

AIMS

To study the theory and practice of digital filters, random processes and pattern processing.

SYLLABUS

Digital Filters and Fourier Transforms (5L)

• Design of Finite Impulse Response (FIR) filters – Window function method
• Design of Infinite Impulse Response (IIR) filters – Bilinear Transformation
• Optimal Design Methods
• Fast Fourier Transform Algorithm

Random processes, optimal signal processing and signal modelling (6L)

• Discrete time random processes - Autocorrelation function, stationarity, power spectrum, ergodicity
• Optimal signal processing - Matched filters and Wiener filters
• Signal Modelling - Autoregressive models, sinusoidal models. Least squares, maximum likelihood, Bayesian parameter estimation.

Pattern Recognition (5L) Prof. Z. Gharamani

• Bayes' decision theory
• Decision boundaries, multivariate Gaussian distributions
• Linear discriminant functions, single layer perceptrons

OBJECTIVES

To become familiar with the fundamental concepts of statistical signal processing, digital filtering and pattern recognition. The student will first study the famous Fast Fourier Transform (FFT) algorithm, which is fundamental to very many of the signal processing operations in practical use today. They will then learn how to design and implement filters (low-pass, band-pass, high-pass) in the digital domain. We will also study the remarkable topic of optimal system design for random systems, leading to the classic Wiener and matched filtering solutions for estimation and detection of signals buried in noise (as used for example in background noise removal for voice recordings and mobile telephony). In the final section of the course students study the fundamental elements of classification and statistical pattern recognition.

ASSESSMENT

By examination

REFERENCES

Please see the Booklist for Part IIA Courses for references for this module.

Last updated: September 2011

teaching-office@eng.cam.ac.uk