Time-Constrained Node Visit Planning for Collaborative UAV-WSN Distributed Applications.

Area of Science:

• Computer Science
• Electrical Engineering
• Robotics

Background:

• Unmanned Aerial Vehicles (UAVs) are increasingly used for data collection in Wireless Sensor Networks (WSNs).
• Current path-planning methods for UAVs in WSNs typically assume data is always ready for retrieval, limiting UAVs to passive data collectors.
• This rigid model overlooks the potential for UAVs to actively engage in computation within the network.

Purpose of the Study:

• To analyze a novel scenario where a UAV actively participates in distributed computation within multiple WSNs.
• To develop and validate a UAV path-planning strategy that integrates computational roles beyond simple data retrieval.
• To enhance network performance by leveraging the UAV's computational capabilities.

Main Methods:

• A new UAV path-planning approach is proposed, enabling the UAV to initiate distributed computation on sensor nodes during its visits.
• The UAV collects computation outcomes on subsequent visits, orchestrating the overall process.
• The method was experimentally validated against existing UAV path-planning techniques.

Main Results:

• The proposed method significantly optimizes total flight time.
• Demonstrated improvements in Average Age of Information (AoI) and Average cluster computation end time.
• Achieved reduced Average data collection time compared to adapted prevalent approaches.

Conclusions:

• UAVs can transition from passive data collectors to active computational entities within WSNs.
• The developed path-planning strategy effectively integrates UAV computation, leading to substantial performance gains.
• This research opens new avenues for intelligent UAV-assisted WSN operations.