Foundations of Resilient Distributed Resource Allocation in Open Networks

Abstract

Networked resource allocation is a fundamental problem in many scientific disciplines and prevalent in numerous civilian and military applications, including spectrum management in tactical wireless networks and multi-robot task allocation for search-and-rescue missions. The goal of this allocation problem is to assign limited resources to a network of agents while optimizing some criteria, for example, allocating wireless spectrum to mobile users with minimal interference. Solving this problem in modern networks is challenging due to a number of factors, such as large-scale and time-varying environments, unexpected failures, physical limitations, and adversaries. These issues result in fundamental challenges of varying numbers of agents and dynamic networks. This project will address these challenges by developing new mathematical foundations and algorithmic techniques to solve dynamic resource allocation problems in open networks with varying numbers of agents. The project will also contribute to education and workforce development by involving students from underrepresented groups in the research. This project will develop a novel distributed online method to decompose the challenging large-scale, time-varying resource allocation problem into a sequence of tractable subproblems. The method advances and integrates several techniques in optimization, online learning, and control to design new scalable, resilient asynchronous distributed solutions for each subproblem. These solutions will then be combined using an innovative surrogate model approach to solve the dynamic allocation problem sequentially. During the course of this project, the proposed research will be applied to two applications- wireless spectrum management and multi-robot task allocation. The expected outcome of this project will be a set of mathematical tools and techniques that can be applied to solve practical problems with autonomous agents acting cooperatively in large-scale dynamic networks. Successful completion of the proposed research activities will result in new effective strategies that will provide improved resource utilization in both civilian and military applications.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 04, 2025
Source ID
FA95502410111

Entities

People

  • Ryan Williams

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • Virginia Tech

Tags

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Distributed Systems and Data Platform Development
  • Operations Research

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • Autonomy
  • Autonomy - Autonomous System Control