A novel data-driven learning method for radar target detection in nonstationary environments
Abstract
Most existing radar algorithms are developed under the assumption that the environment (clutter) is stationary. However, in practice, the characteristics of the clutter can vary enormously depending on the radar-operational scenarios. If unaccounted for, these nonstationary variabilities may drastically hinder the radar performance. Therefore, to overcome such shortcomings, we develop a data-driven method for target detection in nonstationary environments. In this method, the radar dynamically detects changes in the environment and adapts to these changes by learning the new statistical characteristics of the environment and by intelligibly updating its statistical detection algorithm. Specifically, we employ drift detection algorithms to detect changes in the environment; incremental learning, particularly learning under concept drift algorithms, to learn the new statistical characteristics of the environment from the new radar data that become available in batches over a period of time. The newly learned environment characteristics are then integrated in the detection algorithm. Furthermore, we use Monte Carlo simulations to demonstrate that the developed method provides a significant improvement in the detection performance compared with detection techniques that are not aware of the environmental changes.
- Authors:
-
- Univ. of Pittsburgh, Pittsburgh, PA (United States)
- Washington Univ., St. Louis, MO (United States)
- Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
- Publication Date:
- Research Org.:
- Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1253252
- Grant/Contract Number:
- AC05-00OR22725
- Resource Type:
- Accepted Manuscript
- Journal Name:
- IEEE Signal Processing Letters
- Additional Journal Information:
- Journal Volume: 23; Journal Issue: 5; Journal ID: ISSN 1070-9908
- Publisher:
- IEEE Signal Processing Society
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 47 OTHER INSTRUMENTATION; data-driven adaptive radar; cognitive radar; nonstationary environment; incremental learning; active drift learning
Citation Formats
Akcakaya, Murat, Nehorai, Arye, and Sen, Satyabrata. A novel data-driven learning method for radar target detection in nonstationary environments. United States: N. p., 2016.
Web. doi:10.1109/LSP.2016.2553042.
Akcakaya, Murat, Nehorai, Arye, & Sen, Satyabrata. A novel data-driven learning method for radar target detection in nonstationary environments. United States. https://doi.org/10.1109/LSP.2016.2553042
Akcakaya, Murat, Nehorai, Arye, and Sen, Satyabrata. Tue .
"A novel data-driven learning method for radar target detection in nonstationary environments". United States. https://doi.org/10.1109/LSP.2016.2553042. https://www.osti.gov/servlets/purl/1253252.
@article{osti_1253252,
title = {A novel data-driven learning method for radar target detection in nonstationary environments},
author = {Akcakaya, Murat and Nehorai, Arye and Sen, Satyabrata},
abstractNote = {Most existing radar algorithms are developed under the assumption that the environment (clutter) is stationary. However, in practice, the characteristics of the clutter can vary enormously depending on the radar-operational scenarios. If unaccounted for, these nonstationary variabilities may drastically hinder the radar performance. Therefore, to overcome such shortcomings, we develop a data-driven method for target detection in nonstationary environments. In this method, the radar dynamically detects changes in the environment and adapts to these changes by learning the new statistical characteristics of the environment and by intelligibly updating its statistical detection algorithm. Specifically, we employ drift detection algorithms to detect changes in the environment; incremental learning, particularly learning under concept drift algorithms, to learn the new statistical characteristics of the environment from the new radar data that become available in batches over a period of time. The newly learned environment characteristics are then integrated in the detection algorithm. Furthermore, we use Monte Carlo simulations to demonstrate that the developed method provides a significant improvement in the detection performance compared with detection techniques that are not aware of the environmental changes.},
doi = {10.1109/LSP.2016.2553042},
journal = {IEEE Signal Processing Letters},
number = 5,
volume = 23,
place = {United States},
year = {Tue Apr 12 00:00:00 EDT 2016},
month = {Tue Apr 12 00:00:00 EDT 2016}
}
Web of Science