Article Details

Abstract

Aiming at the problem of uncoordinated cloud-edge resource utilization and task allocation in cloud-edge task scheduling of highway monitoring systems, this paper proposes a method (IALG) based on cloud-edge-end collaboration architecture combining improved ant colony algorithm fused with LSTM and LightGBM classification to optimize task allocation and load balancing. Through simulation comparison of multiple methods, this paper comprehensively analyzes three indicators of task completion time, resource utilization rate, and task offloading latency under multiple scenarios. Simulation results show that under four load scenarios of light, normal, heavy, and extreme loads, the task completion time of the system is significantly shortened compared with other methods, the load balance is improved by about 40%, and the task offloading latency is significantly reduced. The IALG algorithm can more effectively shorten task execution time, balance node load, reduce task offloading latency, and demonstrate stronger robustness.

Keywords

cloud-edge-end collaboration; edge computing; Federated Learning; improved ant colony algorithm; task scheduling

References

1. [1] Li Jin, Fan Tengfei, Gao Hongliang, etc Research on Multi core Task Offload Scheduling and Resource Adaptation in edge computing Networks [J]. Ordnance Automation, 2025, 44 (3): 29-34
2. [2] Zhou Xu, Miao Hui, Yang Jing, etc Edge computing Resource Scheduling: History, Architecture, Modeling and Methodology Analysis [J]. Computer Integrated Manufacturing System, 2025, 31 (8): 2695-2726
3. [3] Liu Shumei, An Yisheng, Mu Chen, etc The dependency task scheduling and computation offloading strategy for time delay perception in the Internet of Vehicles [J]. Chinese Journal of Highways, 2024, 37 (9): 221-235
4. [4] Xie Yanyu, Chen Lansun Multi objective task scheduling method based on edge computing [J]. Journal of Jinan University: Natural Science Edition, 2025, 39 (1): 117-122
5. [5] Zhang Jing, Guan Jianfeng, Liu Kexian, etc Edge Computing Network Task Scheduling Algorithm Based on Dynamic Potential Game [J]. Acta Sinica, 2025, 53 (1): 221-237
6. [6] Chen Junren, Guo Yijing Cloud computing task scheduling based on exchange improved particle swarm algorithm [J]. Computer Applications and Software, 2023, 40 (6): 223-228
7. [7] Ren Xiaoqiang, Nie Qingbin, Wang Haoyu, etc Cloud computing task scheduling based on particle swarm optimization and improved ant colony algorithm [J]. Computer Engineering and Design, 2024, 45 (6): 1797-1804
8. [8] Nie Qingbin Cloud computing resource scheduling based on ant colony simulated annealing algorithm [J]. Computer Applications, 2024, 44 (S2): 139-142
9. [9] Chen Jun, Wang Xin, Zeng Hao, etc Online task scheduling for edge heterogeneous clusters based on resource matching [J]. Modern Electronic Technology 2025, 48 (6): 31-38
10. [10] Dong Yumin, Zhang Jing, Xie Lvzuo, etc A review of key issues in edge intelligent computing under cloud edge architecture: computation optimization and computation offloading [J]. Journal of Electronics and Information Technology, 2024, 46 (3): 765-776
11. [11] Tian Zhuojing, Huang Zhenchun, Zhang Yinong A Review of Task Scheduling Methods in Cloud Computing Environments [J]. Computer Engineering and Applications, 2021, 57 (2): 1-11
12. [12] BONADIO A，CHITI F，FANTACCI R. Performance Analysis of an Edge Computing SaaS System for Mobile Users [J].IEEE Transactions onVehicular Technology，2019，69（2）：2049-2057.
13. [13] GANG Y N，LIU X S，TONG D H，et al. Non -invasive Power Load Monitoring Method Based on Cloud Edge Collaboration [C]//2020 4th International Workshop on Advances in Energy Science and Environment Engineering.Hangzhou：IOP Publishing Ltd，2020，512：190-195.
14. [14] Wu Teng, Wang Zeyu Research on Cloud Edge Collaboration Technology in Edge IoT [J]. China Broadband, 2023, 19 (11): 148-150
15. [15] Zhang Shouli, Liu Chen Research on Service based Cloud Edge Collaborative Flow Data Processing Architecture [J]. Journal of Shandong Agricultural University: Natural Science Edition, 2024, 55 (3): 385-395
16. [16] WANG J Y，PAN J L，ESPOSITO F，et al. Edge Cloud Offloading Algorithms: Issues, Methods and Perspectives [J].ACM Computing Surveys(CSUR)，2019，52（1）：1-23.
17. [17] LI J Y，YUAN H T，MA Y F，et al. Low-Latency and Energy-Efficient Task Scheduling for End-Edge-Cloud Collaborative Computing [C]//2024 IEEE International Conference on Systems, Man, and Cybernetics (SMC).Kuching：IEEE，2024：611-616.
18. [18] Dai Ning, Xu Kaixin, Hu Xudong, etc Information Collection and Monitoring System for Loom Equipment under Cloud Edge Collaboration [J]. Journal of Textile Science, 2024, 45 (8): 215-224
19. [19] Shi Jianfeng, Chen Xinyang, Li Baolong Research on Task Offloading and Resource Allocation Algorithms in Cloud Edge Collaborative Computing for the Internet of Things [J]. Journal of Electronics and Information Technology, 2025, 47 (2): 458-469
20. [20] Cui Shuangshuang, Wu Jian, Wang Hongzhi, etc Multi modal data modeling technology and its application for cloud edge collaboration [J]. Journal of Software, 2024, 35 (3): 1154-1172
21. [21] GAO X Q，HU Y M，SHAO Y Z，et al. Hierarchical Dynamic Resource Allocation for Computation Offloading in LEO Satellite Networks [J].IEEE Internet ofThings Journal，2024，11（11）：19470-19484.

Funding

Research and development of expressway new intelligent electromechanical edge computing MCU product based on autonomous trusted operating system (YCIC-YF-2023-02); Risk Identification, Assessment, and Visualization System for Unstable Slopes Along Typical Highways in Yunnan Province (JKYZLX-2022-23)

About the authors

Sun Hongxian (1984-), male, Han ethnicity, from Kunming, Yunnan, senior engineer, master's degree, research interests include transport information digitalization and intelligent transportation; Corresponding author: Ni Wei (1987-), male, Han ethnicity, from Kunming, Yunnan, engineer, bachelor's degree, research interests include transport information digitalization and intelligent transportation.