Date of Award
Thesis - Restricted
Master of Science (MS)
Electrical and Computer Engineering
Heinen, James A.
Akbay, Kunter S.
Floating-point digital signal processors aid in implementing real-time digital signal processing algorithms in hardware. Several different hardware architectures will be examined for their merits in implementing digital signal processing systems with digital signal processors. Specifically, an examination of a single processor architecture, a pipelined architecture, a parallel architecture, a multipipelined architecture, and a neural architecture will be made. Software architectures play an important role in the implementation of a digital signal processing system. Software development environments, low-level firmware, diagnostics, the run-time kernel, and high-level languages will be introduced and discussed for their importance in building a good digital signal processing system. Applications for digital signal processors are vast. A brief overview of applications will introduce finite impulse response (FIR) filters, infinite impulse response (IIR) filters, fast Fourier transforms (FFTs), Kalman filters, robotic control, adaptive algorithms, and array processors. Next, a method for designing a digital signal processing system · to be implemented in real-time is proposed. First, a system is expressed in some manner which allows the precedence and parallel nature of operations to be determined. This is followed by the allocation of the desired number of digital signal processors. Additionally, some graphical methods of representing these algorithms will be introduced. Finally, some illustrative examples are provided utilizing the proposed design methods.
Frigo, Fred J., "Implementing Real-Time Digital Signal Processing Algorithms in Hardware" (1989). Master's Theses (1922-2009) Access restricted to Marquette Campus. 4569.