Observe
Use available sensor positions, received measurements, and localization outputs to describe the current sensing state.
SUPPORTING TECHNOLOGY
Autonomous sensor tasking.
Research in choosing where cooperative RF sensors should move to improve the next localization measurement.
THE CONTROL PROBLEM
Make each new measurement more useful.
Sensor tasking connects a sensing objective to a movement decision. For cooperative geolocation, the objective is to improve the collective estimate of an emitter location as the array geometry changes.
Select
Evaluate candidate sensor movements using a learned policy and the geometry objectives of the experiment.
Reassess
Collect another snapshot and evaluate how the estimate changes before choosing the next action.
TWO RELATED RESEARCH THREADS
DQN control and decentralized learning.
The flagship DQN work evaluates adaptive geometry control within a cooperative localization system. A separate Asilomar study explores decentralized Q-learning with consensus policy selection. These are related research approaches with distinct architectures and evaluations.
- Geometry Adaptive Deep Q-Network for UAV-Based Emitter Localization in Cluttered RF Environments 2026 IEEE Aerospace Conference · 2026
- Decentralized Q-Learning for Emitter Localization with Consensus Policy Selection 59th Asilomar Conference on Signals, Systems, and Computers · 2025
TRANSITION QUESTIONS
From a learned policy to an integrated system.
A practical implementation needs a defined sensor interface, synchronized measurements, communication assumptions, motion constraints, and platform-level testing. These requirements shape the next stage of technology development and evaluation.
CONNECT WITH KRONMERIC
Explore a research partnership.
Discuss cooperative RF sensing, algorithm evaluation, or a technology development opportunity.