Optimization Based Approaches to Network Analytics

Abstract

The U.S. Air Force vision of net-centric warfare requires development and deployment of large-scale complex systems that will encompass and interconnect various manned and unmanned weapon platforms and combat units. Efficient and reliable operation of the resulting complex, heterogeneous, multilayered networks (which may include C3 networks, sensory networks, ad-hoc wireless networks connecting battlefield units, etc) will require a deep understanding of the properties and behavior of such networks, particularly in the context of incomplete, conflicting, or uncertain operational data. The objective of this research project was to develop theoretical and computational tools for network analytics, comprising a collection of optimization-based frameworks and methods for comprehensive analysis of complex, heterogeneous, multilayered networks involving imperfect/flawed/uncertain data, common in military operations. The underlying theme of this project has been the development and study of predictive metrics and characterizations of network performance that are specifically defined and constructed through solutions of optimization problems. The conducted research consisted in a fusion of graph-theoretical approaches with the methodologies of discrete/combinatorial, stochastic, and robust optimization, with the focus on obtaining globally optimal solutions for new data-driven network analytic models.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 14, 2022
Accession Number
AD1230522

Entities

People

  • Pavlo Krokhmal

Organizations

  • University of Arizona

Tags

Readers

  • Neural Network Machine Learning.
  • Systems Analysis and Design

Technology Areas

  • Autonomy
  • Autonomy - Autonomous System Control
  • Autonomy - UAVs
  • Fully Networked C3
  • Fully Networked C3 - Command and Control