Mathematical methods and techniques in signal processing [E9 252(3:0) – Fall 2013]

Instructor:
Shayan G. Srinivasa,

[Tuesday, Thursday 11:30 am – 1 pm, DESE auditorium]


Pre-requisities:

Undergraduate signals and systems/DSP course.


Course Syllabus:

  • Review of basic signals, systems and signal space: Review of 1-D signals and systems, review of random signals, multi-dimensional signals, review of vector spaces, inner product spaces, orthogonal projections and related concepts.
  • Basics of multi-rate signal processing: sampling, decimation and interpolation, sampling rate conversion (integer and rational sampling rates), oversampled processing (A/D and D/A conversion), and introduction to filter banks.
  • Signal representation: Transform theory and methods (FFT and variations, KLT), other transform methods.
  • Statistical signal modeling: The least squares method, Pade’s approximation, Prony’s method, Shanks’ method, iterative pre-filtering, all-pole modeling and linear prediction, autocorrelation and covariance methods, FIR least squares inverse filter design, applications and examples.
  • Inverse problems (signal reconstruction): underdetermined least squares, pseudo-inverse (SVD), min-norm solutions, regularized methods, reconstruction from projections, iterative methods such as projection onto convex sets, expectation-maximization and simulated annealing.

Reference Books:

  • Moon & Stirling, Mathematical Methods and Algorithms for Signal Processing, Prentice Hall, 2000. (required)
  • Monson Hayes, Statistical Digital Signal Processing and Modeling, John Wiley and Sons, 1996. (optional)
  • Class notes

Grading Policy: 

Policy #1 Policy #2
Homeworks : 15%

Exam #1 : 15%

Exam #2 : 20%

Project : 15%

Final Exam : 35%

Homeworks : 15%

Exam #1 : 15%

Exam #2 : 20%

Project : 20%

Final Exam : 30%

The final grade is max (Policy#1,Policy#2) which ever works best for the student.


Homeworks


Exams


Project