Numerical and Symbolic Algorithms for Application Specific Signal Processing
Final rept. 1 May 93-31 Dec 97
MASSACHUSETTS INST OF TECH CAMBRIDGE RESEARCH LAB OF ELECTRONICS
Pagination or Media Count:
Our research under the RASSP program focussed on the development of new signal processing algorithms to exploit the advantages and opportunities of rapid prototyping for signal processing systems. Our work was carried out in the Research Laboratory of Electronics at MIT with close collaboration and interaction with industrial partners including Lockheed Sanders. Our primary accomplishments were in several categories. We developed and explored new classes of signals and algorithms related to chaotic behavior in nonlinear systems. We also developed and explored the use of fractal signals for communications. A third new nonlinear signal class that we explored exploits the solution behavior of certain nonlinear wave equations. A significant part of our effort was directed at the exploration of algorithm-based fault tolerant architectures for signal processing and for the use of approximate processing techniques which can be exploited in the contexts of low-power signal processing architectures and in other aspects of processor resource management. We also explored opportunities to exploit multirate signal processing. In the following section we briefly summarize our results in each of these areas. Section 3 contains a bibliography of the journal articles, book chapters, and conference papers in which the details of our work is presented.
- Operations Research
- Electricity and Magnetism