Passive EM Analysis Using Array Processing and Machine Learning

Abstract

Funds are provided to examine changes in Radio Frequency recieved phase and power due to refractivity gradients in the coastal atmosphere by utilizing modeling and simulation to demonstrate the sensitivity of the complex-valued field to deterministic and random components of the environment. We will explore the limits of our ability to observe the complex field using a system based on commercially available software defined radios. Finally, we will explore how machine learning may be utilized to go beyond the limitations of linear/quadratic methods. The potential knowledge yield of this research includes: Marginal benefit of complex data in (EM) environmental estimation; Use of wider range of emitter signals for passive environmental characterization; Relation of array parameters to capability for environmental sensing; Improvements to inversion performance by moving past current implementations such as (linear/quadratic) of Bayesian paradigm.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 05, 2021
Source ID
N000142112587

Entities

People

  • Peter Gerstoft

Organizations

  • Office of Naval Research
  • United States Navy
  • University of California, San Diego

Tags

Readers

  • Neural Network Machine Learning.
  • Ocean-Atmosphere Mesoscale Modeling, Data Assimilation, and Flux Boundary Layers
  • Sensor Fusion and Tracking Systems.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks