A compressive sensing-driven data collection scheme for UAV-assisted wireless sensor networks.

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Unmanned Aerial Vehicles (UAVs) are increasingly used for data collection in sensor networks.
• Energy constraints in remote areas challenge the long-term operation of UAVs and sensor nodes (SNs).
• Efficient energy utilization is critical for sustainable UAV-assisted sensor networks.

Purpose of the Study:

• To propose a novel sparse compressive sensing (CS) sampling framework for UAV-assisted data collection.
• To develop an energy optimization scheme that minimizes overall system energy consumption.
• To enhance the long-term and stable operation of energy-constrained UAVs and SNs.

Main Methods:

• A sparse compressive sensing (CS) sampling framework was developed.
• An energy optimization scheme was designed, decomposing the problem into node selection and joint optimization sub-problems.
• The CS-based Node Selection (CSNS) and Two-Stage Joint Optimization (TSJO) algorithms were created, utilizing successive convex approximation (SCA), variable substitution, and relaxation.

Main Results:

• The CSNS algorithm optimizes UAV flight distance while balancing network energy consumption under CS reconstruction accuracy constraints.
• The TSJO algorithm jointly optimizes SN scheduling, UAV flight duration, and UAV trajectory.
• Simulation results validated the effectiveness of both CSNS and TSJO algorithms.

Conclusions:

• The proposed UAV-assisted data collection scheme significantly reduces energy consumption compared to benchmark methods.
• The developed CSNS and TSJO algorithms effectively address energy efficiency challenges in UAV-assisted sensor networks.
• This framework enables more sustainable and long-term operation of sensor networks utilizing UAVs.