In this experiment, students learn the basics of real-time signal processing for embedded systems. Examples of embedded devices are mobile phone, digital camera and vehicle's anti-lock braking system (ABS). Signal processing algorithms used on those devices should be adapted to hardware constraints.
The heart of the experiment involves the conversion of an algorithm from a simulation environment to a heterogeneous real-time system that includes a DSP core while implementing an example application of acoustic echo cancellation. The experiment consists of two parts. In the first part, students perform audio enhancement by echo cancellation using adaptive filtering. It includes theory of adaptive filtering practiced in MATLAB and in C. In the second part, embedded implementation is performed using the OMAP system-on-chip from Texas Instruments. OMAP contains a DSP core for signal processing tasks and an ARM core for general-purpose tasks. Implementation in the experiment uses both cores and is optimized by the students, taking into account embedded considerations such as fixed-point representation.