边缘计算下的多无人机野外协同作业机制研究
|
摘要
无人机编队在野外或灾后作业过程中因外界通信链路不佳且能量有限,面临调度难、续航短的问题.本文针对该问题提出了一种基于边缘计算的多机协同互助作业方案,所有计算任务都在最边缘的设备(无人机)上协同完成.将无人机节点建模为Jackson网络,在任务卸载调度时综合考虑通信成本、各节点的实时计算能力和剩余能量,继而提出了Oracle式(O)、主动集中式(PC)、主动分布式(PD)、反馈分布式(RD)四种调度算法,并设计模拟器进行仿真实验,将几种算法与非协作方法一起进行效率和能耗的综合比较.实验结果显示,协作方法明显优于非协作方法,当外部任务到达率较高时,PC可以在性能和能耗间取得最佳平衡,当任务到达率较低时RD表现更佳.
In the field or post disaster operation,UAV formation is confronted with poor scheduling and short mileage due to poor communication links and limited energy. In this paper,a multi UAV cooperative mutual assistance scheme based on edge computing is proposed.All of the computing tasks are processed on the most marginal devices(unmanned aerial vehicles). The UAV node is modeled as Jackson network,and the four kinds of Oracle(O),proactive centralised(PC),proactive distributed(PD),and reactive distributed(RD)are put forward by considering the communication cost,the real-time computing power and the residual energy of each node in the task unloading and scheduling. The simulator is designed to carry out simulation experiments. The efficiency and energy consumption of several algorithms are compared with those of non-cooperative methods. The experimental results show that the cooperative method is obviously better than the non-cooperative method. When the arrival rate of the external task is high,the proactive centralized(PC)can achieve the best balance between the performance and the energy consumption,and the reactive distributed(RD)performance is better when the task arrival rate is low.
In the field or post disaster operation,UAV formation is confronted with poor scheduling and short mileage due to poor communication links and limited energy. In this paper,a multi UAV cooperative mutual assistance scheme based on edge computing is proposed.All of the computing tasks are processed on the most marginal devices(unmanned aerial vehicles). The UAV node is modeled as Jackson network,and the four kinds of Oracle(O),proactive centralised(PC),proactive distributed(PD),and reactive distributed(RD)are put forward by considering the communication cost,the real-time computing power and the residual energy of each node in the task unloading and scheduling. The simulator is designed to carry out simulation experiments. The efficiency and energy consumption of several algorithms are compared with those of non-cooperative methods. The experimental results show that the cooperative method is obviously better than the non-cooperative method. When the arrival rate of the external task is high,the proactive centralized(PC)can achieve the best balance between the performance and the energy consumption,and the reactive distributed(RD)performance is better when the task arrival rate is low.
引文
[1] Fang Kun,Re Le,Yu Yun-long,et al. Cross-layer method for reliable information transmission in UAV Ad hoc netw orks based on priority scheduling[J]. Journal of Chinese Computer Systems,2017,38(5):1028-1033.
[2] Zhu Qing,Han Hui-peng,Yu Jie,et al. Multi-objective optimization scheduling method for UAV resources in emergency surveying and mapping[J]. Geomatics and Information Science of Wuhan University,2017,42(11):1608-1615.
[3] Shi Wei-song,Sun Hui,Cao Jie,et al. Edge computing—an emerging computing model for the internet of everything era[J]. Journal of Computer Research and Development,2017,54(5):907-924.
[4] Kumar K,Yung Hsiang Lu. Cloud computing for mobile users:can offloading computation save energy?[J]. Computer,2010,43(4):51-56.
[5] Li Ru,Lu Ya-fei,Hou Zhong-xi. A model of mission planning for cooperative UAVs[C]. Proceedings of Control and Decision Conference(CCDC),IEEE,2015:67-72.
[6] Gu J,Su T,Wang Q,et al. Multiple moving targets surveillance based on a cooperative netw ork for multi-UAV[J]. IEEE Communications M agazine,2018,56(4):82-89.
[7] Budaev D,Amelin K,Voschuk G,et al. Real-time task scheduling for multi-agent control system of UAV's group based on netw orkcentric technology[C]. Proceedings of Control,Decision and Information Technologies(CoDIT),2016 International Conference on,IEEE,2016:378-381.
[8] Simi S,Kurup R,Rao S. Distributed task allocation and coordination scheme for a multi-uav sensor netw ork[C]. Proceedings of Wireless and Optical Communications Netw orks(WOCN),2013Tenth International Conference on IEEE,2013:1-5.
[9] Zhou Yue-zhi,Zhang Di. Near-end cloud computing:opportunities and challenges in the post-cloud computing era[J]. Chinese Journal of Computers,2018,41(25):1-23.
[10] Stewart W J. Probability,Markov chains,queues,and simulation:the mathematical basis of performance modeling[M]. Princeton University Press,2009.
[11] Vilaplana J,Solsona F,Rius J,et al. A queuing theory model for cloud computing[J]. Journal of Supercomputing,2014,69(1):492-507.
[12] Chen Qing-ping. High speed ad hoc network routing protocol in air[D]. Xi'an:Xidian University,2011.
[13] Meskar E,Todd T D,Zhao D,et al. Energy aware offloading for competing users on a shared communication channel[J]. IEEE Transactions on M obile Computing,2016,16(1):87-96.
[14] Wu H,Sun Y,Wolter K. Analysis of the energy-response time tradeoff for delayed mobile cloud offloading[J]. ACM SIGM ETRICS Performance Evaluation Review,2015,43(2):33-35.
[15] Zivkovic Z. Improved adaptive Gaussian mixture model for background subtraction[C]. Pattern Recognition(ICPR)2004,Proceedings of the 17th International Conference on,IEEE,2004,2:28-31.
[16] Jung W,Kang C,Yoon C,et al. DevScope:a nonintrusive and online pow er analysis tool for smartphone hardw are components[C]. Proceedings of the Eighth IEEE/ACM/IFIP International Conference on Hardw are/Softw are Codesign and System Synthesis,2012:353-362.
[17] Likamwa R,Priyantha B,Philipose M,et al. Energy characterization and optimization of image sensing tow ard continuous mobile vision[C]. Proceedings of the International Conference on M obile Systems,Applications,and Services,ACM,2013:69-82.
[18] Dargie W. A Stochastic model for estimating the power consumption of a processor[J]. IEEE Transactions on Computers,2015,64(5):1311-1322.
[19] Bai F,Helmy A. A survey of mobility models in wireless adhoc netw orks[C]. Wireless Ad Hoc and Sensor Netw orks,Springer,2004:1-30.
[1]方堃,茹乐,于云龙,等.优先级调度下的无人机自组网数据可靠传输跨层方法[J].小型微型计算机系统,2017,38(5):1028-1033.
[2]朱庆,韩会鹏,于杰,等.应急测绘无人机资源多目标优化调度方法[J].武汉大学学报(信息科学版),2017,42(11):1608-1615.
[3]施巍松,孙辉,曹杰,等.边缘计算:万物互联时代新型计算模型[J].计算机研究与发展,2017,54(5):907-924.
[9]周悦芝,张迪.近端云计算:后云计算时代的机遇与挑战[J].计算机学报,2018,41(25):1-23.
[12]陈清平.空中高速自组织网络路由协议[D].西安:西安电子科技大学,2011.
[2] Zhu Qing,Han Hui-peng,Yu Jie,et al. Multi-objective optimization scheduling method for UAV resources in emergency surveying and mapping[J]. Geomatics and Information Science of Wuhan University,2017,42(11):1608-1615.
[3] Shi Wei-song,Sun Hui,Cao Jie,et al. Edge computing—an emerging computing model for the internet of everything era[J]. Journal of Computer Research and Development,2017,54(5):907-924.
[4] Kumar K,Yung Hsiang Lu. Cloud computing for mobile users:can offloading computation save energy?[J]. Computer,2010,43(4):51-56.
[5] Li Ru,Lu Ya-fei,Hou Zhong-xi. A model of mission planning for cooperative UAVs[C]. Proceedings of Control and Decision Conference(CCDC),IEEE,2015:67-72.
[6] Gu J,Su T,Wang Q,et al. Multiple moving targets surveillance based on a cooperative netw ork for multi-UAV[J]. IEEE Communications M agazine,2018,56(4):82-89.
[7] Budaev D,Amelin K,Voschuk G,et al. Real-time task scheduling for multi-agent control system of UAV's group based on netw orkcentric technology[C]. Proceedings of Control,Decision and Information Technologies(CoDIT),2016 International Conference on,IEEE,2016:378-381.
[8] Simi S,Kurup R,Rao S. Distributed task allocation and coordination scheme for a multi-uav sensor netw ork[C]. Proceedings of Wireless and Optical Communications Netw orks(WOCN),2013Tenth International Conference on IEEE,2013:1-5.
[9] Zhou Yue-zhi,Zhang Di. Near-end cloud computing:opportunities and challenges in the post-cloud computing era[J]. Chinese Journal of Computers,2018,41(25):1-23.
[10] Stewart W J. Probability,Markov chains,queues,and simulation:the mathematical basis of performance modeling[M]. Princeton University Press,2009.
[11] Vilaplana J,Solsona F,Rius J,et al. A queuing theory model for cloud computing[J]. Journal of Supercomputing,2014,69(1):492-507.
[12] Chen Qing-ping. High speed ad hoc network routing protocol in air[D]. Xi'an:Xidian University,2011.
[13] Meskar E,Todd T D,Zhao D,et al. Energy aware offloading for competing users on a shared communication channel[J]. IEEE Transactions on M obile Computing,2016,16(1):87-96.
[14] Wu H,Sun Y,Wolter K. Analysis of the energy-response time tradeoff for delayed mobile cloud offloading[J]. ACM SIGM ETRICS Performance Evaluation Review,2015,43(2):33-35.
[15] Zivkovic Z. Improved adaptive Gaussian mixture model for background subtraction[C]. Pattern Recognition(ICPR)2004,Proceedings of the 17th International Conference on,IEEE,2004,2:28-31.
[16] Jung W,Kang C,Yoon C,et al. DevScope:a nonintrusive and online pow er analysis tool for smartphone hardw are components[C]. Proceedings of the Eighth IEEE/ACM/IFIP International Conference on Hardw are/Softw are Codesign and System Synthesis,2012:353-362.
[17] Likamwa R,Priyantha B,Philipose M,et al. Energy characterization and optimization of image sensing tow ard continuous mobile vision[C]. Proceedings of the International Conference on M obile Systems,Applications,and Services,ACM,2013:69-82.
[18] Dargie W. A Stochastic model for estimating the power consumption of a processor[J]. IEEE Transactions on Computers,2015,64(5):1311-1322.
[19] Bai F,Helmy A. A survey of mobility models in wireless adhoc netw orks[C]. Wireless Ad Hoc and Sensor Netw orks,Springer,2004:1-30.
[1]方堃,茹乐,于云龙,等.优先级调度下的无人机自组网数据可靠传输跨层方法[J].小型微型计算机系统,2017,38(5):1028-1033.
[2]朱庆,韩会鹏,于杰,等.应急测绘无人机资源多目标优化调度方法[J].武汉大学学报(信息科学版),2017,42(11):1608-1615.
[3]施巍松,孙辉,曹杰,等.边缘计算:万物互联时代新型计算模型[J].计算机研究与发展,2017,54(5):907-924.
[9]周悦芝,张迪.近端云计算:后云计算时代的机遇与挑战[J].计算机学报,2018,41(25):1-23.
[12]陈清平.空中高速自组织网络路由协议[D].西安:西安电子科技大学,2011.