Multi-Objective Simulated Annealing for Efficient Task Allocation in UAV-Assisted Edge Computing for Smart City Traffic Management
Smart city traffic management relies increasingly on UAV-assisted edge computing systems to process real-time data and make informed decisions. A critical challenge in these systems is the efficient allocation of computational tasks across available edge computing resources. While existing technolog...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10870050/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Smart city traffic management relies increasingly on UAV-assisted edge computing systems to process real-time data and make informed decisions. A critical challenge in these systems is the efficient allocation of computational tasks across available edge computing resources. While existing technologies provide solutions for data collection (UAVs), processing (computer vision), and control (reinforcement learning), the integration and resource optimization of these components remains a significant challenge. We propose a multi-objective simulated annealing (MOSA) algorithm for optimizing task allocation in edge computing systems, focusing on three key objectives: minimizing active computational nodes, optimizing energy distribution, and reducing execution time. We compared the MOSA algorithm with uniform random allocation, greedy algorithm, and single-objective simulated annealing (SOSA) methods under both standard and peak load conditions. The peak load scenario tested system performance under significantly increased computational demands and reduced resource availability. Our evaluation focused on three key metrics: the number of active nodes, energy distribution efficiency, and task execution time. The proposed MOSA algorithm demonstrated superior resource utilization under standard conditions and maintained robust performance during peak loads, showing significant improvements over baseline methods in all metrics. Results show that MOSA effectively balances multiple objectives while adapting to varying operational demands. It consistently outperformed comparison methods in minimizing active nodes while maintaining competitive performance in energy distribution and execution time. The framework demonstrated particular strength in maintaining efficiency under significantly increased computational loads, offering a robust solution for task allocation in edge computing systems. While some limitations exist in real-world applications, this work provides a strong foundation for optimizing resource utilization in smart city systems that integrate multiple computational tasks. |
|---|---|
| ISSN: | 2169-3536 |