Southwest Research Institute (SwRI) develops cognitive electronic warfare (CogEW) capabilities optimized for artificial intelligence (AII) and deep learning on novel hardware platforms such as neuromorphic processors and advanced FPGAs. These systems enable real-time observation and analysis of broad frequency ranges, the generation of adaptive, novel responses to emerging threats, and the ability to continuously evolve with new and dynamic signal environments through transformative architectures that enhance decision speed, efficiency and mission effectiveness.
Radio Frequency Machine Learning (RFML)
SwRI is advancing cognitive electronic warfare by moving beyond traditional decision-support–level methods that rely on symbolic engagement models and PDW-based emitter behavior prediction. While waveform-level RFML has seen far less development, it is essential for real-time parameter estimation, classification and prediction directly from raw measurements. Because waveform-level techniques are highly sensitive to environmental and hardware variations, SwRI is focusing on executing these capabilities on real hardware and integrating them into operational systems to achieve reliable, mission-ready performance.
Image
SwRI engineer David Brown discusses the complexities of integrating RF data analytics, big data adaptive sampling and physics-inspired neural networks (PINNs) into cognitive electronic warfare implementations.
Advanced Data Processing & Analysis
SwRI is pushing boundaries on algorithms, hardware technology and applications of RFML using cutting-edge hardware and tools across several critical domains.
Image
Adaptive sampling can help reduce dimensionality, which leads to faster processing and throughput in cognitive electronic warfare systems.
RFML Applications
- RF data analytics
- Big data adaptive sampling
- Physics-inspired neural networks (PINNs)
Hardware & Tools
- Spikes, sparsity and static-suppression
- Electra-MA direct sampling transceivers
Critical Domains
- Autonomous engagement
- Distributed multi-agent cognition
- Constrained optimization
|
Events & Training |
|
|
Course: Cognitive Electronic Warfare: An Artificial Intelligence (AI) Approach |
Life Cycle Industry Days (LCID) & Wright Dialogue with Industry (WDI) |
Image
USAF Taps SwRI for CogEW Algorithms
SwRI awarded $6.4 million to explore detect and respond to unknown enemy radar threats in real time.