Content

Many problems in digital signal processing can be solved much more easily in the frequency domain than in the time domain. Therefore we begin with a repetition of the Fourier transform. In order to process time continuous signals in a digital computer they have to be sampled. This leads us to the well known sampling theorem and the reconstruction of analog signals from samples using low pass filtering. The discrete and fast Fourier transform are used to implement digital filters efficiently. Parallel to the lectures a system is developed to transmit two signals over a common channel in different frequency bands using modulation.

Prerequisites

Course material

Demos
You can download the following Java programs and start them with double click or java -jar program.jar from the command line. All you need is a free Java SDK installation.

Homework
I will upload homework every week here. Some of the exercises are milestones for the project. Try to do them regularly in order to avoid too much workload at the deadline of the project.

Grading
Part of the lecture is a project, which has to be accomplished in a team with maximum 3 persons. Project presentations are scheduled in January. Deadline for submission of a written report (8 pages) is December 31. Please attach the check list.

Literature

Fourier transform, sampling, aliasing, signal reconstruction

FFT, fast convolution