Future ESM Systems and the Potential for Neural Processing,
Abstract
The projected radar electromagnetic environments (eme) for the 1990's and beyond include: higher pulse densities (several million pps); frequencies to 40 GHz and higher; stable, jittered, staggered and pseudo-random pulse repetition intervals (PRIs) with multiple frequencies; spread spectrum techniques; multiple agile radar beams an multi-mode missile seekers. Electronic Support Measures (ESM) concerns the passive detection and identification of radar signals. Thus, an ESM system which can measure such signal characteristics will most likely flood its main processor with information to such an extent that it may not be able to cope. In addition, missing pulses and receiver/processor shadowing times may lead to degraded input data to the processor. A number of likely solutions exist ranging from special purpose hardware to new processing techniques. In the past few years, neural networks have shifted from being primarily a research technology to active use in wide-ranging defence applications. This paper will indicate the likely applicability of a neural processing approach to a range of ESM functions together with results from some preliminary proof-of-concept investigations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1991
- Accession Number
- ADP006334
Entities
People
- Arthur G. Self
- Gregory Bourassa
Organizations
- AGARD