Event-Train Restoration Via Backpropagation Neural Networks
Abstract
This project is investigating backpropagation neural networks for specific applications in passive electronic warfare involving restoration of deinterleaved event trains to their original broadcast form. This is different from traditional bit-error detection/correction which relies on a prior knowledge of what the original bit stream looked liked. In electronic warfare it is unlikely that such prior knowledge will be available. Results of this research can be applied to 3 major problem areas: 1) pulse-train restoration, 2) communications signal compression, and 3) data compression. (EDC)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1989
- Accession Number
- ADA216308
Entities
People
- Peter G. Raeth
Organizations
- Wright Laboratory