车辆边缘网络中基于多参数MDP模型的动态服务迁移策略

doi:10.11959/j.issn.1000-436x.2020012

摘要/Abstract

摘要：

为解决车辆移动及边缘服务器有限服务范围造成的服务中断问题，为车辆边缘网络提出一种基于多参数马尔可夫决策过程的动态服务迁移算法。通过构造包含时延、带宽、服务器处理能力及车辆运动信息的多参数MDP 收益函数，弥补了单纯基于距离进行服务迁移方案的不足；不再使用单一迁移目标服务器，结合车辆运动及时延限制构造候选服务器集合，基于Bellman方程表示的长期收益值进行迁移决策；利用历史数据进行权重计算及数据更新，提高了算法对动态环境的适应能力。仿真结果表明，所提算法降低了服务时延、数据分组丢失率及服务迁移次数。

关键词: 边缘计算, 车辆边缘网络, 动态服务迁移, MDP

Abstract:

To handle with the service interruption caused by vehicles’ mobility and limited service coverage of edge servers,a dynamic service migration algorithm based on multi-parameters Markov decision process (MDP) model was put forward for vehicular edge network,which was called as dynamic service migration algorithm based on multiple parameter (DSMMP).Combining delay,bandwidth,server capacity with vehicle motion information,DSMMP constructed a multi-parameters MDP revenue function to remedy the deficiency of distance-based schemes.By using vehicle motion and delay constraints,a candidate server set with several candidate servers was defined,and migration decision through long-term Bellman revenue values was made.In order to improve the dynamic adaptability of the proposed algorithm,the weight values were calculated and updated by leveraging historical information.Simulation results show that our strategy has a good performance in terms of delay,packet loss ratio and service migration times.

Key words: edge computing, vehicular edge network, dynamic service migration, MDP

中图分类号:

TP393

郭辉,芮兰兰,高志鹏. 车辆边缘网络中基于多参数MDP模型的动态服务迁移策略[J]. 通信学报, 2020, 41(1): 1-14.

Hui GUO,Lanlan RUI,Zhipeng GAO. Dynamic service migration strategy based on MDP model with multiple parameter in vehicular edge network[J]. Journal on Communications, 2020, 41(1): 1-14.

图/表 22

表1

图1

图2

图3

表2

表3

图4

图5

图6

图7

表4

历史数据矩阵 A m "

编号	$R_{D}^{m} (s, a)$	$R_{B}^{m} (s, a)$	$R_{L}^{m} (s, a)$	$R_{M}^{m} (s, a)$
X₁	A₁₁	A₁₂	A₁₃	A₁₄
X₂	A₂₁	A₂₂	A₂₃	A₂₄
X₃	A₃₁	A₃₂	A₃₃	A₃₄

表4

表5

图8

图9

图10

图11

图12

图13

图14

图15

图16

图17

参考文献 30

[1]	ABICHANDANI P , FLIGOR W , FROMM E . A cloud enabled virtual reality based pedagogical ecosystem for wind energy education[C]// IEEE Frontiers in Education Conference. IEEE, 2014: 1-7.
[2]	KHAN A A , REHMANI M H , REISSLEIN M . Cognitive radio for smart grids:survey of architectures,spectrum sensing mechanisms,and networking protocols[J]. IEEE Communication Survey ＆ Tutorials, 2016,18(1): 860-898.
[3]	REHMANI R H , KANTARCI M E , RACHEDI A ,et al. Cognitive radio based smart grid:the future of the traditional electrical grid[J]. Ad Hoc Networks, 2016,41: 1-4.
[4]	REHMANI M H , KANTARCI M E , RACHEDI A ,et al. Smart grids:a hub of interdisciplinary research[J]. IEEE Access, 2015(3): 3114-3118.
[5]	施巍松, 刘芳, 孙辉 ,等. 边缘计算[M]. 北京: 科学出版社, 2017.
	SHI W S , LIU F , SUN H ,et al. Edge computing[M]. Beijing: Science PressPress, 2017.
[6]	IEEE Communication Society . Standard for adoption of open fog reference architecture for fog computing[S].,(2018-08-01)[2019-09-30]. IEEE Standard,(2018-08-01)[2019-09-30].
[7]	SHI W S , CAO J , ZHANG Q ,et al. Edge computing:vision and challenges[J]. IEEE Internet of Things Journal, 2016,3(5): 637-646.
[8]	Cisco. Cisco visual networking index:global mobile data traffic forecast update,2016-2021[R].[2019-09-30].
[9]	ABBAS N , ZHANG Y , TAHERKORDI A ,et al. Mobile edge computing:a survey[J]. IEEE Internet of Things Journal, 2018,5(1): 450-465.
[10]	WANG S , ZHANG X , ZHANG Y ,et al. A survey on mobile edge networks:convergence of computing,caching and communications[J]. IEEE Access, 2017(5): 6757-6779.
[11]	EJAZ A , HUSAIN R M . Mobile edge computing:opportunities,solutions,and challenges[J]. Future Generation Computer Systems, 2017,70: 59-63.
[12]	LI H X , SHOU G C , HU Y H ,et al. Mobile edge computing:progress and challenges[C]// 2016 4th IEEE International Conference on Mobile Cloud Computing,Services,and Engineering. IEEE, 2016: 83-84.
[13]	PATEL M , HU Y C , HEDE P ,et al. Mobile-edge computing Introductory technical white paper[R]. Mobile-Edge Computing (MEC) Industry Initiative,(2014-09)[2019-09-30]
[14]	俞一帆, 任春明, 陈思仁 . 5G移动边缘计算[M]. 北京: 人民邮电出版社, 2017.
	YU Y F , REN C M , CHEN S R . 5G mobile edge computing[M]. Beijing: Posts and Telecom PressPress, 2017.
[15]	ETSI MEC . MEC in an enterprise setting:a solution outline[R].(2018-09)[2019-09-30].
[16]	WANG S G , XU J L , ZHANG N ,et al. A survey on service migration in mobile edge computing[J]. IEEE Access, 2018(6): 23511-23528.
[17]	TALEB T , KSENTINI A , FRANGOUDIS P . Follow-me cloud:when cloud services follow mobile users[J]. IEEE Transactions on Cloud Computing, 2018,7(2): 369-382.
[18]	TALEB T , KSENTINI A . An analytical model for follow me cloud[C]// 2013 IEEE Global Communications Conference. IEEE, 2013: 1291-1296.
[19]	WANG S Q , URGAONKAR R , HE T ,et al. Mobility-induced service migration in mobile micro-clouds[C]// 2014 IEEE Military Communications Conference. IEEE, 2014: 835-840.
[20]	SHARMA R , KUMAR S , TRIVEDI M C . Mobile cloud computing:a needed shift from cloud to mobile cloud[C]// 2013 5th International Conference and Computational Intelligence and Communication Networks. 2013: 533-539.
[21]	WANG S Q , URGAONKAR R , ZAFER M ,et al. Dynamic service migration in mobile edge-clouds[C]// 2015 IFIP Networking Conference. IFIP, 2015: 1-9.
[22]	URGAONKAR R , WANG S Q , HE T ,et al. Dynamic service migration and workload scheduling in edge-clouds[C]// 33rd International Symposium on Computer Performance,Modeling,Measurements,and Evaluation / IFIP WG 7.3 Performance. IFIP, 2015: 205-228.
[23]	NEELY M J . Stochastic network optimization with application to communication and queueing systems[M]. Sam Rafael: Morgan and Claypool PublishersPress, 2010.
[24]	ZHANG W Y , HU Y , ZHANG Y Y ,et al. SEGUE:quality of service aware edge cloud service migration[C]// 2016 IEEE International Conference on Cloud Computing Technology and Science. IEEE, 2016: 344-351.
[25]	FEI Z M , BHATTACHARJEE S , ZEGURA E W ,et al. A novel server selection technique for improving the response time of a replicated service[C]// INFOCOM’98.Seventeenth Annual Joint Conference of the IEEE Computer and Communications Societies. IEEE, 1998: 783-791.
[26]	ZHAO D , YANG T , JIN Y H ,et al. A service migration strategy based on multiple attribute decision in mobile edge computing[C]// 2017 IEEE 17th International Conference on Communication Technology. IEEE, 2017: 986-990.
[27]	GAO Z G , CHEN D J , CAI S B ,et al. OptDynLim:an optimal algorithm for the one-dimensional RSU deployment problem with non- uniform profit density[J]. IEEE Transactions on Industrial Informatics, 2019,15(2): 1052-1061.
[28]	GAO Z G , CHEN D J , CAI S B ,et al. Optimal and greedy algorithms for the one-dimensional RSU deployment problem with new model[J]. IEEE Transactions on Vehicular Technology, 2018,67(8): 7643-7657.
[29]	RAYCHAUDHURI D , SESKAR I , OTT M ,et al. Overview of the ORBIT radio grid testbed for evaluation of next-generation wireless network protocols[C]// IEEE Wireless Communications and Networking Conference. IEEE, 2005: 1664-1669.
[30]	PIORKOWSKI M , SARAFIJANOVIC D N , GROSSGLAUSER M . CRAWDAD dataset epfl/mobility[R].(2009-02-24)[2019-09-30].

方案	设计	优点	缺点
文献[17]方案	利用基于跳数的 MDP 状态函数对蜂窝网络中的用户运动进行聚类，迭代求解	提高了对用户行为预测的准确度	可扩展性低
文献[21]方案	利用“常数+指数”的形式具体化MDP开销函数，改进策略迭代以获得更优策略	实现了MDP函数具体化	计算复杂度高、动态适应性弱
文献[22]方案	利用Lyapunov优化及MDP的解耦特性将约束的MDP问题转化为简单的确定性问题，获得高效求解	简化MDP模型，提高了求解效率	忽略了环境的动态性，可靠性差
文献[24]方案	基于服务器响应时间变化检测性能冲突并进行迁移决策，选择一个累计QoS收益最高的ES作为目标ES	利用时延冲突主动进行服务迁移，与被动方案相比更及时、可靠	忽略了动态环境特点，缺乏监测数据准确性的保障机制
文献[26]方案	考虑了每个虚拟机的多个属性因素并建立相应的决策矩阵，通过决策矩阵来决策迁移	充分考虑了能量、时延等开销的影响，实用性较高	忽略了网络环境的动态性，缺乏参数的实时更新机制

R_label	timer/ms	ES_label
Request₁	0.5	ES₁
Request₃	1	ES₂
#	#	#

V_label	R_label	hop
V₁	Request₁	0
	Request₄
V₂	Request₃	3
	Request₇
#	#	#

参数	值
平台	ORBIT
区域	40 km×40 km
ES服务覆盖半径/m	2 100
ES服务重叠区域/m	204
ES个数	10×10
BCS个数	1
车辆个数	250
车辆速度/(m.s^-1)	10~20（随机分配）
每次仿真运行时间/s	100
请求时间间隔/s	1
请求数据大小/MB	1~64
带宽/MB	100~150（随机分配）
时延/ms	1~50（随机分配）
α、β、δ、ε	初始均为0.25
γ	0.9
状态检测时隙t/s	1
T_w	初始为2t