Dynamic Resource Allocation for a Sensor Network

Abstract

We describe a project that analyzes the performance of a sensor network trying to detect enemy agents in an urban environment. We simulate a network of units that can detect different enemy types using a variety of sensors. Our work aims to address key issues about networked systems, including the effectiveness of cooperative strategies, the balance of multiple constraints, and the tradeoffs between resource allocation and performance. We make use of evolutionary computing algorithms to search the space of possible configurations, carrying out parameter sweeps to test how various dimensions of the design space impact performance over a range of scenarios. Our results provide some quantitative insights into the performance of networked systems.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 2005
Accession Number
ADA464105

Entities

People

  • Eric Bonabeau
  • Iavor Trifonov
  • Martin C. Martin
  • Paolo Gaudiano

Tags

Communities of Interest

  • Energy and Power Technologies
  • Human Systems
  • Materials and Manufacturing Processes
  • Sensors
  • Weapons Technologies

DTIC Thesaurus Topics

  • Accuracy
  • Agent-Based Simulations
  • Algorithms
  • Command And Control
  • Complex Adaptive Systems
  • Complex Systems
  • Computations
  • Control Systems
  • Detectors
  • Evolutionary Algorithms
  • Genetic Algorithms
  • Mathematical Analysis
  • Network Centric Warfare
  • Networks
  • Sensor Networks
  • Simulations
  • Vehicle Borne Improvised Explosive Devices

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Computational Modeling and Simulation
  • Sensor Fusion and Tracking Systems.

Technology Areas

  • Space