Performance of an Adaptive Matched Filter Using the Griffiths Algorithm
Abstract
This thesis presents a CW and noise jamming analysis of an adaptive matched filter that (1) uses the Griffiths algorithm and (2) has a pseudonoise sequence as an input. The analysis is conducted over several jamming powers, frequencies, and phases. The Griffiths adaptive matched filter is shown to converge for raised cosine pulses that experience no distortion, quadratic delay distortion, and cubic delay distortion. The Griffiths adaptive matched filter diverges for pulses that experience linear delay distortion even though the the convergence rate constant is within limits. Throughout the analysis the P-vector is determined apriori and held constant. The Griffiths filter is shown to converge for CW jamming and noise jamming. Noise jamming is shown to be more effective in the higher power ranges. A comparison is made between the Griffiths adaptive matched filter and an adaptive matched filter that uses the LMS algorithm. The degradation is performance of the Griffiths filter compared to an LMS filter that uses a stored reference is calculated for several selected runs. The actual computer programs used are presented in the Appendix.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1988
- Accession Number
- ADA202578
Entities
People
- Peter D. Pasko
Organizations
- Air Force Institute of Technology