Conference proceeding
Disentangling high energy chorus elements against structured background interference in the Van Allen radiation belts using braid manifolds
2020 54th Asilomar Conference on Signals, Systems, and Computers, pp.73-77
11/01/2020
DOI: 10.1109/IEEECONF51394.2020.9443336
Abstract
This paper presents a geometric signal processing technique and related feature representation using braid manifolds to disentangle buried patterns of interest against structured background interference. Specifically, we will explore the theoretical underpinnings of how algebraic geometry, and in particular, braid groups, may be harnessed with traditional signal processing to autonomously create robust training sets and domain-interpretable feature dictionaries for different machine learning architectures. The target application aimed in this work is to autonomously detect chorus elements in the Earth's Van Allen radiation belts. Understanding chorus wave generation is fundamental to space weather prediction of high-energy "killer electrons" that are well-known to cause long-lasting if not fatal damage to space instrumentation, including defense telecommunication satellites, in the Van Allen radiation belts. Beyond chorus element detection, due to the fundamentally mathematical nature of braided feature engineering, our method is broadly applicable across any remote sensing paradigm, e.g. passive sonar sensing of the marine environment, active sensing of buried underwater mines, acoustic sensing of the dynamic shallow water channel in communication scenarios, etc. Theory and techniques developed will be validated over publicly available magnetospheric data from NASA's Van Allen Radiation Belts (also known as Van Allen Probes) mission.
Details
- Title: Subtitle
- Disentangling high energy chorus elements against structured background interference in the Van Allen radiation belts using braid manifolds
- Creators
- Ananya Sen Gupta - University of IowaRyan A McCarthy - University of IowaKawther Rouabhi - University of IowaCraig Kletzing - University of IowaIvar Christopher - University of Iowa
- Resource Type
- Conference proceeding
- Publication Details
- 2020 54th Asilomar Conference on Signals, Systems, and Computers, pp.73-77
- DOI
- 10.1109/IEEECONF51394.2020.9443336
- eISSN
- 2576-2303
- Publisher
- IEEE
- Language
- English
- Date published
- 11/01/2020
- Academic Unit
- Electrical and Computer Engineering; Physics and Astronomy
- Record Identifier
- 9984197533802771
Metrics
22 Record Views