Reference Number: ARL-SRP-00034

Project Description

Distributed networks and edge sensor arrays operate under limited communication bandwidth, power, and computing capacity. Continuously processing and transmitting sensor data can consume scarce resources, while sharing too little can degrade estimation accuracy. How can collaborating agents determine which measurements are worth acquiring, processing, and communicating? This project will investigate resource-allocation policies guided by the expected reduction in decision-relevant uncertainty. The objective is to determine when selectively sensing, processing, and sharing information can maintain estimation accuracy while reducing resource use.

Working with technical mentors, the student will build or extend a modular simulation in which multiple agents use different sensors to track a moving source under communication and computing constraints. The student will implement baseline tracking filters and fixed-schedule policies, develop uncertainty-guided alternatives, and compare their performance across varying resource budgets, communication dropouts, and node failures. Evaluation will examine tracking error, communication volume, computational cost, and whether reported uncertainty accurately reflects estimation errors. A key outcome will be understanding when adaptive coordination improves performance and when it introduces delay, redundancy, or overconfidence.

Anticipated tools include Python, scientific computing libraries, and Bayesian estimation methods such as extended Kalman filtering. Expected deliverables include a simulation testbed, benchmark results comparing adaptive and fixed-schedule policies under comparable resource constraints, and a technical report. Results may also support a conference publication. Helpful background includes Python programming experience and familiarity with one or more of the following: distributed systems, estimation or control theory, machine learning, and networking.

Helpful background includes Python programming experience and familiarity with one or more of the following: distributed systems, estimation or control theory, machine learning, and networking.

Technical Skills

• Advanced Sensing
• Artificial Intelligence/Machine Learning
• Computer Modeling/Simulation
• Computer Science
• Data Science
• Electrical Engineering
• Mathematics