CERN, Geneva
Field Programmable Gate Arrays (FPGAs) have become a central enabling technology for the design of fast digital signal processing systems. This tutorial starts with an introduction to digital signal processing and a comparison with analog techniques. We then treat the problem of choosing between the two key technologies for digital systems: Digital Signal Processors (DSPs) and FPGAs. Once the advantages of FPGAs for very demanding systems have been laid out, we go on with a survey of digital design techniques of general nature, followed by tips and tricks more directly applicable to FPGA implementations. Digital signal processing in FPGAs typically uses a fixed-point number representation. We explain how different fixed-point arithmetic operations can be implemented, and the trade-offs regarding speed, silicon area and precision. Finally, all the concepts are applied to a set of examples in beam instrumentation.
