IITM BS Digital Signal Processing (EE3101): Syllabus and Tips

Digital Signal Processing (EE3101) is a diploma course in the IITM BS in Electronic Systems. It covers signals in discrete time: sampling, the discrete-time Fourier transform (DTFT), the z-transform, the frequency response of digital filters, and the DFT with the FFT. You need Signals and Systems (EE2101) first. The syllabus gives it 5 credits, while the handbook gives 4.

The syllabus names Prof. R. David Koilpillai of the Department of Electrical Engineering, IIT Madras, as the instructor.

The credit mismatch

The Sep 2025 syllabus lists 4-0-1-5: 4 lecture credits and 1 practical credit, 5 in total. The handbook's course table lists DSP at 4 credits. The syllabus does not describe what the practical credit involves, and the handbook lists no separate DSP lab. Check the credits on your registration page.

What you learn in DSP

The syllabus spends two weeks on each topic.

• Weeks 1 and 2: review of discrete-time signals and systems. Basic signals, system properties such as stability and causality, systems described by difference equations, impulse response and convolution.
• Weeks 3 and 4: sampling. How the spectrum of a sampled signal relates to its DTFT, true frequency versus digital frequency, rebuilding a signal with sinc interpolation, aliasing, oversampling, and processing analog signals digitally.
• Weeks 5 to 8: the DTFT and the z-transform. The DTFT, its inverse and properties. Then the z-transform, its region of convergence, three ways to invert it, pole-zero plots, and what causality and stability mean for the region of convergence.
• Weeks 9 and 10: frequency response. Magnitude and phase from a pole-zero plot, simple filters like notch, comb, averaging and allpass, group delay, minimum phase systems, and the four types of linear phase FIR filters.
• Weeks 11 and 12: the DFT. The DFT as samples of the DTFT, its matrix form, circular shift and circular convolution, zero padding, and the FFT algorithms.

The suggested books are Discrete-Time Signal Processing by Oppenheim and Schafer, and DSP books by Tarun K. Rawat, by Proakis and Manolakis, and by Sanjit Mitra.

As a diploma course, it needs worked-out assignment sheets. The syllabus does not give a grading formula.

Who finds it hard and how to prepare

DSP moves fast through a lot of theory. The z-transform with its region of convergence, and the phase and group delay ideas in weeks 9 and 10, need the most practice.

• Make sure convolution and the Fourier transform from Signals and Systems are solid. Weeks 1 and 2 are a review, and they go quickly.
• For pole-zero questions, practise the geometric method on filters with just one or two poles and zeros first.
• Work a 4 point DFT fully by hand once. It makes the matrix form and the FFT much easier to follow.
• Check your hand answers with a plotting tool. If you know some Python, plotting a frequency response takes a few lines.
• Revise the circulant matrix and DFT week of Maths II before weeks 11 and 12.

What to take before and after

Before: Signals and Systems (EE2101), and ideally Math for Electronics II (MA2101), which covers the DFT from the matrix side. After or with it: the Signal Processing Project (EE4999), which lists both Signals and Systems and DSP as corequisites. See ES diploma courses for the full list.

Common questions

Is DSP 4 credits or 5 credits?

The syllabus says 5 (4-0-1-5). The handbook says 4. The difference is the 1 practical credit in the syllabus. Check your registration page, or ask support which value applies to you.

Can I take DSP and Analog Electronic Systems in the same term?

Yes. Both need only Signals and Systems. The medium and fastest study plans both put DSP, Analog Electronic Systems and Digital System Design in the same term.

How is DSP different from Signals and Systems?

Signals and Systems works mostly in continuous time, with the Fourier and Laplace transforms. DSP moves to discrete time and adds sampling, the z-transform, digital filter design ideas and the DFT.