Page 70 - 《软件学报》2020年第9期

软件学报 ISSN 1000-9825, CODEN RUXUEW E-mail: jos@iscas.ac.cn
Journal of Software,2020,31(9):2691−2708 [doi: 10.13328/j.cnki.jos.005938] http://www.jos.org.cn
©中国科学院软件研究所版权所有. Tel: +86-10-62562563

∗
面向边缘计算的目标追踪应用部署策略研究

张 展, 张宪琦, 左德承, 付国栋

(哈尔滨工业大学 计算机科学与技术学院,黑龙江 哈尔滨 150001)
通讯作者: 张展, E-mail: zhangzhan@hit.edu.cn

摘 要: 目标追踪算法虽已在诸多领域得到广泛应用,然而由于实时性和功耗问题,使得基于深度学习模型的算
法难以在移动终端设备上部署应用.结合边缘计算技术,从应用部署优化的角度,对目标追踪算法在移动设备上的部
署策略进行研究.通过对目标追踪应用特点、移动设备特性以及边缘云网络架构的分析,提出一种面向边缘计算的
目标追踪应用部署策略.通过任务分割策略,将目标追踪应用的计算任务合理卸载至边缘云,并利用信息融合策略对
计算结果进行分析融合;此外,利用运动检测,进一步降低终端节点的计算压力和功耗.通过对不同部署策略进行对
比实验,结果表明:相比计算任务本地计算,该部署策略明显降低了任务响应时间;相比完全卸载至边缘云,该部署策
略降低了相同计算任务的处理时间.
关键词: 目标追踪;边缘计算;资源分配;深度学习;移动计算
中图法分类号: TP391

中文引用格式: 张展,张宪琦,左德承,付国栋.面向边缘计算的目标追踪应用部署策略研究.软件学报,2020,31(9):2691−2708.
http://www.jos.org.cn/1000-9825/5938.htm
英文引用格式: Zhang Z, Zhang XQ, Zuo DC, Fu GD. Research on target tracking application deployment strategy for edge
computing. Ruan Jian Xue Bao/Journal of Software, 2020,31(9):2691−2708 (in Chinese). http://www.jos.org.cn/1000-9825/
5938.htm
Research on Target Tracking Application Deployment Strategy for Edge Computing

ZHANG Zhan, ZHANG Xian-Qi, ZUO De-Cheng, FU Guo-Dong
(School of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001, China)

Abstract: Target tracking algorithm has been widely used in many fields. However, due to the problems of real-time and power
consumption, it is difficult to deploy the algorithm based on deep learning model on mobile terminal devices. This work studies the
deployment strategy of target tracking algorithm on mobile devices from the perspective of application deployment optimization
combined with edge computing technology. A deployment strategy of target tracking application oriented to edge computing is proposed
based on the analysis of device characteristics and edge cloud network architecture. The computing task of target tracking application is
reasonably unloaded to edge cloud by task segmentation strategy and the computing results are analyzed and fused by the information
fusion strategy. In addition, a motion detection scheme is proposed to further reduce the computing pressure and power consumption of
terminal nodes The experimental results show that compared with local computing, the deployment strategy significantly reduces the
response time of the task, and compared with completely uninstalling to the edge cloud, the deployment strategy reduces the processing
time of the same computing task.
Key words: target tracking; edge computing; resource allocation; deep learning; mobile computing

∗ 基金项目: 国家高技术研究发展计划(863)(2013AA01A215)
Foundation item: National High Technology Research and Development Program of China (863) (2013AA01A215)
本文由“智能嵌入式系统”专题特约编辑王泉教授、吴中海教授、陈仪香教授、苗启广教授推荐.
收稿时间: 2019-06-29; 修改时间: 2019-08-18; 采用时间: 2019-11-02; jos 在线出版时间: 2020-01-13
CNKI 网络优先出版: 2020-01-14 11:26:54, http://kns.cnki.net/kcms/detail/11.2560.TP.20200114.1126.020.html