This project establishes a teaching laboratory for undergraduate instruction in digital signal processing (DSP). The laboratory gives students practical, hands-on training in DSP concepts and techniques using modem hardware and software design tools that complement the more theoretical approach of a lecture course in DSP. The laboratory uses PC-type workstations with add-on DSP processing boards and a complement software to provide both non-real-time and real-time DSP environments for algorithm development and testing. The new laboratory course consists of a dozen laboratory exercises, about one per week during the semester, each emphasizing a particular topic in DSP (for example, convolution, sampling, filter design, and Fourier transformation). In the laboratory, students program non-real-time simulations of common DSP procedures using MATLAB and compare the results of these simulations with theoretical predictions. They then implement these same DSP procedures in real time using a hardware/software system that allows them to graphically program a DSP coprocessor card. Students have the opportunity to visualize and hear the results of applying real-time DSP procedures to real-world signals such as speech and music. The proposed laboratory augments and enhances offerings in the School of Engineering and the Department of Computer Science in digital signal processing, control and robotics, computer music, and multimedia.
