意图驱动的自动驾驶网络技术

doi:10.19665/j.issn1001-2400.2022.04.008

摘要/Abstract

摘要：

由于终端用户、垂直行业和内容提供商等对未来网络的灵活性和智能性等提出更高要求,所以网络服务多样化,网络结构复杂化,迫切需求网络规划、管理、运维和优化等自主化,最终实现网络意图服务意图的高度自治。为降低网络管理的复杂度,提升网络自优化的程度,需探索和利用自动化、意图驱动网络、人工智能和策略自动生成等新技术实现自动驾驶网络。自动驾驶网络把意图驱动网络作为当前网络演进的愿景,依托其意图转译技术、策略生成验证技术等实现基于意图的网络高效运营管理,并结合人工智能技术减少人工运维的开销,提高网络管理优化的效率。总结了自动驾驶网络的研究背景;明晰了自动驾驶网络的定义及优势;进而,提出了新型的自动驾驶网络的网络管控架构、实现流程和关键技术;最后设计典型自动驾驶网络的应用实例,以证明实现自动驾驶网络愿景的可行性并明确自动驾驶网络的发展方向。自动驾驶网络通过为用户提供更灵活、更高效的网络管控能力,高效实现用户需求,提升用户体验。

关键词: 意图驱动网络, 网络自治, 人工智能, 闭环控制

Abstract:

End-users,vertical industries,content providers,etc.present higher requirements for future network flexibility and intelligence.The diversification of network services and the complexity of network architectures urgently require the autonomy of network planning,management,operation and optimization,and ultimately achieve a high degree of autonomy in network intent and service intent.To reduce the complexity of network management and enhance the degree of network self-optimization,new technologies such as automation,intent-driven networks,artificial intelligence and automatic policy generation need to be explored and utilized to realize autonomous driving networks.The autonomous driving network takes the intent-driven network as the vision of the current network evolution,and relies on its intent translation technology and strategy generation verification technology to realize the efficient operation and management of the network based on the intent.In addition,it combines artificial intelligence technology to reduce the cost of manual operation and maintenance and improve the efficiency of network management optimization.This paper summarizes the research background of autonomous driving networks,clarifies the definition and advantages of autonomous driving networks and then,proposes a new autonomous driving network control architecture,implementation process and key technologies.Finally,a typical autonomous driving network application example is designed to demonstrate the feasibility of realizing the vision of autonomous driving networks and clarify the development direction of autonomous driving networks.The autonomous driving network provides users with more flexible and efficient network management and control capabilities,effectively realizes user demands and improves user experience.

Key words: Intent-driven Networks, network autonomy, artificial intelligence, closed-loop control

中图分类号:

TN915.07

冷常发,杨春刚,彭瑶. 意图驱动的自动驾驶网络技术[J]. 西安电子科技大学学报, 2022, 49(4): 60-70.

LENG Changfa,YANG Chungang,PENG Yao. Intent-driven autonomous driving networking technology[J]. Journal of Xidian University, 2022, 49(4): 60-70.

图/表 7

表1

图1

图2

图3

表2

图4

表3

参考文献 28

[1]	PANG L, YANG C, CHEN D, et al. A Survey on Intent-Driven Networks[J]. IEEE Access, 2020, 8:22862-22873. doi: 10.1109/ACCESS.2020.2969208
[2]	THUPAE R, ISONG B, GASELA N, et al. Software Defined Wireless Sensor Networks Management and Security Challenges:A Review[C]// IECON 2018 44th Annual Conference of the IEEE Industrial Electronics Society.Piscataway:IEEE, 2018:4736-4741.
[3]	HEORHIADI V, CHANDRASEKARAN S, REITER M K, et al. Intent-Driven Composition of Resource-Management SDN Applications[C]// Proceedings of the 14th International Conference on Emerging Networking Experiments and Technologies. New York: ACM, 2018:86-97.
[4]	WANG T H, CHEN Y C, HUANG Y T, et al. IPDAC:An Integrated IP Address Management Framework for Telecommunication Management Networks[C]// 2019 20th Asia-Pacific Network Operations and Management Symposium (APNOMS).Piscataway:IEEE, 2019:1-4.
[5]	KALMBACH P, ZERWAS J, BABARCZI P, et al. Empowering Self-Driving Networks[C]// Proceedings of the Afternoon Workshop on Self-Driving Networks. New York: ACM, 2018:8-14.
[6]	MADANAPALLI S C, GHARAKHEILI H H, SIVARAMAN V. AssistingDelay and Bandwidth Sensitive Applications in A Self-Driving Network[C]// Proceedings of the 2019 Workshop on Network Meets AI & ML. New York: ACM, 2019:64-69.
[7]	ZERWAS J, KALMBACH P, HENKEL L, et al. Netboa:Self-Driving Network Benchmarking[C]// Proceedings of the 2019 Workshop on Network Meets AI & ML. New York: ACM, 2019:8-14.
[8]	YAQOOB T, USAMA M, QADIR J, et al. On Analyzing Self-Driving Networks:A Systems Thinking Approach[C]// Proceedings of the Afternoon Workshop on Self-Driving Networks. New York: ACM, 2018:1-7.
[9]	HUANG H, DING S, ZHAO L, et al. Real-Time Fault Detection for IIoT Facilities Using GBRBM-Based DNN[J]. IEEE Internet of Things Journal, 2019, 7(7):5713-5722. doi: 10.1109/JIOT.2019.2948396
[10]	程强, 刘姿杉. 电信网络智能化发展现状与未来展望[J]. 信息通信技术与政策, 2020, 46(9):16-22.
	CHENG Qiang, LIU Zishan. Intelligent Telecommunication Networks:Current Status and Future Prospects[J]. Information and Communications Technology and Policy, 2020, 46(9):16-22.
[11]	OKUYAMA T, GONSALVES T, UPADHAY J. AutonomousDriving System Based on Deep Q Learnig[C]// 2018 International Conference on Intelligent Autonomous Systems (ICoIAS).Piscataway:IEEE, 2018:201-205.
[12]	ZHAO Z, SCHILLER E, KALOGEITON E, et al. AutonomicCommunications in Software-Driven Networks[J]. IEEE Journal on Selected Areas in Communications, 2017, 35(11):2431-2445. doi: 10.1109/JSAC.2017.2760354
[13]	GACANIN H. AutonomousWireless Systems with Artificial Intelligence:A Knowledge Management Perspective[J]. IEEE Vehicular Technology Magazine, 2019, 14(3):51-59. doi: 10.1109/MVT.2019.2920162
[14]	ZEYDAN E, TURK Y. Recent Advances in Intent-Based Networking:A Survey[C]// 2020 IEEE 91st Vehicular Technology Conference (VTC2020-Spring).Piscataway:IEEE, 2020:1-5.
[15]	UJCICH B E, BATES A, SANDERS W H. Provenance forIntent-Based Networking[C]// 2020 6th IEEE Conference on Network Softwarization (NetSoft).Piscataway:IEEE, 2020:195-199.
[16]	AREZOUMAND S, DZEPAROSKA K, BANNAZADEH H, et al. MD-IDN:Multi-Domain Intent-Driven Networking in Software-Defined Infrastructures[C]// 2017 13th International Conference on Network and Service Management (CNSM).Piscataway:IEEE, 2017:1-7.
[17]	VEIGA A P. Applications of Artificial Intelligence to Network Security (2018)[J/OL]. [2018-03-27]. https://arxiv.org/abs/1803.09992.
[18]	LILLICRAP T P, HUNT J J, PRITZEL A, et al. Continuous Control with Deep Reinforcement Learning (2015)[J/OL]. [2015-09-09]. https://arxiv.org/abs/1509.02971.
[19]	HYUN J, HONG J W K. Knowledge-Defined Networking Using In-Band Network Telemetry[C]// 2017 19th Asia-Pacific Network Operations and Management Symposium (APNOMS).Piscataway:IEEE, 2017:54-57.
[20]	MARQUES J A, LUIZELLI M C, DA COSTA FILHO R I T, et al. An Optimization-Based Approach for Efficient Network Monitoring Using In-Band Network Telemetry[J]. Journal of Internet Services and Applications, 2019, 10(1):1-20. doi: 10.1186/s13174-018-0098-z
[21]	KIM C, SIVARAMAN A, KATTA N, et al. In-band Network Telemetry via Programmable Dataplanes[J]. Sosr, 2015, 2:2-3.
[22]	TAN L, SU W, ZHANG Z, et al. In-Band Network Telemetry:A Survey[J]. Computer Networks, 2021, 186:107763. doi: 10.1016/j.comnet.2020.107763
[23]	MARQUES J A, LUIZELLI M C, FILHO R, et al. An Optimization-Based Approach for Efficient Network Monitoring Using In-Band Network Telemetry[J]. Journal of Internet Services and Applications, 2019, 10:1-20. doi: 10.1186/s13174-018-0098-z
[24]	CHEN Y, CHEN C, CHANG H, et al. Constructing ECA Rule for IoT Application through a Novel S2RG Process:The Exemplary ECA Rules for Smarter Energy Applications[C]// 2016 International Computer Symposium (ICS).Piscataway:IEEE, 2016:549-554.
[25]	KARIMI L, JOSHI J. An Unsupervised Learning Based Approach for Mining Attribute Based Access Control Policies[C]// 2018 IEEE International Conference on Big Data (Big Data).Piscataway:IEEE, 2018:1427-1436.
[26]	ISAZADEHA A, PEDRYCZW, MAHAN F. ECA Rule Learning in Dynamic Environments[J]. Expert Systems with Applications, 2014, 41(17):7847-7857. doi: 10.1016/j.eswa.2014.06.028
[27]	YANG H, ZHAN K, YAO Q, et al. Intent Defined Optical Network with Artificial Intelligence-Based Automated Operation and Maintenance[J]. Science China Information Sciences, 2020, 63(6):1-12.
[28]	JACOBS A S, PFITSCHER R J, FERREIRA R A. Refining Network Intents for Self-Driving Networks[C]// Proceedings of the Afternoon Workshop on Self-Driving Networks. New York: ACM, 2018:15-21.

领域	学术界	产业界
定义	自动测量、分析和控制自身的网络	跨域、跨层级融合的智慧网络通信架构
	意图深度挖掘	网络高度自治自愈
特点	网络状态全局感知	主动识别网络意图
	网络配置实时优化	快速处理网络故障
	博弈决策	智能管理
关键技术	实时遥测	故障恢复
	优化控制	闭环优化

分类	常见协议和方案
主动测量	PING、Traceroute、Iperf、IPMP、OWAMP、TWAMP、MPLS L/DM、Pingmesh等^[22]
被动测量	Netflow、sFlow、IPFIX、PSAMP等^[23]
混合测量	Reactive Measurement、In-band Measurement、AM-PM等^[24]

阶段	网络管理系统	网络功能
人工级	分立工具	实现通信功能完成基本连接
推荐级	跨域分析知识图谱	构建网络拓扑管理数据回传
控制级	智能决策	路径质量感知多路由调度
自治级	多种环境接入	数字孪生网络