智能化卫星互联网运维与管理：现状与机遇

doi:10.11959/j.issn.2096-8930.2021046

摘要/Abstract

摘要：

在多层异构的卫星互联网中，时变的网络拓扑、动态的链路质量、不稳定的卫星节点、海量涌现的高维数据给网络的运维与管理带来巨大挑战。人工智能技术由于具备从海量历史数据中学习或从与动态环境的交互反馈中学习的能力，被认为是应对上述挑战的有效方案。在网络资源管理、业务流量管理和网络状态管理3个方面重点调研人工智能技术在卫星网络运维与管理领域的研究现状，在此基础上深入分析现有方法存在的不足，并讨论未来值得关注的研究方向。

关键词: 卫星互联网, 人工智能, 网络运维与管理, 资源分配, 异常检测

Abstract:

In multi-layer heterogeneous satellite internet, the time-varying topology, dynamic channel quality, unstable satellites and emerging high-dimensional data present daunting challenges to its maintenance and management system.Thanks to its ability to learn from massive historical data or interaction with dynamic environment, artificial intelligence has been considered to be a promising solution to meet the aforementioned challenge.In this paper, we reviewed the research status of intelligent satellite internet maintenance and management technology on the following three aspects, network resource management, traff c fl ow management and network state management.Based on the above introduction, we further discussed the limitation of existing research and discussed promising future research direction of this fi eld as well.

Key words: satellite internet, artifi cial intelligence, network maintenance and management, resource allocation, anomaly detection

中图分类号:

TP393

唐斯琪, 潘志松, 胡谷雨, 吴炀, 李云波. 智能化卫星互联网运维与管理：现状与机遇[J]. 天地一体化信息网络, 2021, 2(4): 75-83.

Siqi TANG, Zhisong PAN, Guyu HU, Yang WU, Yunbo LI. Intelligent Satellite Internet Maintenance and Management:Progress and Opportunities[J]. Space-Integrated-Ground Information Networks, 2021, 2(4): 75-83.

图/表 5

图1

表1

表2

图2

图3

参考文献 22

[1]	吴巍 . 天地一体化信息网络发展综述[J]. 天地一体化信息网络, 2020,1(1): 1-16.
	WU W . Survey on the development of space-integrated-ground information network[J]. Space-Integrated-Ground Information Networks, 2020(1): 1-16.
[2]	周笛, 盛敏, 郝琪 ,等. 巨型星座系统的网络运维与资源管控技术[J]. 天地一体化信息网络, 2020,1(1): 26-35.
	ZHOU D , SHENG M , HAO Q ,et al. Network operation,maintenance and resource management in mega constellation SystemFull text replacement[J]. Space-Integrated-Ground Information Networks, 2020,1(1): 26-35.
[3]	LIU S J , HU X , WANG W D . Deep reinforcement learning based dynamic channel allocation algorithm in multi-beam satellite systems[J]. IEEE Access, 2018,6: 15733-15742.
[4]	HU X , LIAO X L , LIU Z J ,et al. Multi-agent deep reinforcement learning-based flexible satellite payload for mobile terminals[J]. IEEE Transactions on Vehicular Technology, 2020,69(9): 9849-9865.
[5]	HU X , LIU S J , WANG Y P ,et al. Deep reinforcement learningbased beam Hopping algorithm in multibeam satellite systems[J]. IET Communications, 2019,13(16): 2485-2491.
[6]	QIU C , YAO H P , YU F R ,et al. Deep Q-learning aided networking,caching,and computing resources allocation in software-defined satellite-terrestrial networks[J]. IEEE Transactions on Vehicular Technology, 2019,68(6): 5871-5883.
[7]	ZHOU C H , WU W , HE H L ,et al. Deep reinforcement learning for delay-oriented IoT task scheduling in SAGIN[J]. IEEE Transactions on Wireless Communications, 2021,20(2): 911-925.
[8]	FERREIRA P V R , PAFFENROTH R , WYGLINSKI A M ,et al. Reinforcement learning for satellite communications:from LEO to deep space operations[J]. IEEE Communications Magazine, 2019,57(5): 70-75.
[9]	魏伍, 张更新, 吕晶 ,等. 一种用于卫星互联网的业务预测方法[J]. 通信技术, 2015,48(11): 1285-1289.
	WEI W , ZHANG G X , LYU J ,et al. Traffc prediction method for satellite network[J]. Communications Technology, 2015,48(11): 1285-1289.
[10]	潘成胜, 王羽夫, 杨力 . 基于改进LSTM算法的天地一体化信息网络流量预测[J]. 天地一体化信息网络, 2020,1(2): 57-65.
	PAN C S , WANG Y F , YANG L . Traffic prediction of spaceintegrated-ground information network based on improved LSTM algorithm[J]. Space-Integrated-Ground Information Networks, 2020,1(2): 57-65.
[11]	LI N , HU L , DENG Z L ,et al. Research on GRU neural network satellite traffc prediction based on transfer learning[J]. Wireless Personal Communications, 2021,118(1): 815-827.
[12]	YANG L , GU X X , SHI H F . A noval satellite network traffic prediction method based on GCN-GRU[C]// Proceedings of 2020 International Conference on Wireless Communications and Signal Processing (WCSP). Piscataway:IEEE Press, 2020: 718-723.
[13]	YANG T T , LI J B , FENG H L ,et al. A novel transmission scheduling based on deep reinforcement learning in softwaredefined maritime communication networks[J]. IEEE Transactions on Cognitive Communications and Networking, 2019,5(4): 1155-1166.
[14]	QIN Z Y , YAO H P , MAI T L . Traffic optimization in satellites communications:a multi-agent reinforcement learning approach[C]// Proceedings of 2020 International Wireless Communications and Mobile Computing (IWCMC). Piscataway:IEEE Press, 2020: 269-273.
[15]	YAIRI T , TAKEISHI N , ODA T ,et al. A data-driven health monitoring method for satellite housekeeping data based on probabilistic clustering and dimensionality reduction[J]. IEEE Transactions on Aerospace and Electronic Systems, 2017,53(3): 1384-1401.
[16]	PAN D W , SONG Z , NIE L Q ,et al. Satellite telemetry data anomaly detection using Bi-LSTM prediction based model[C]// Proceedings of 2020 IEEE International Instrumentation and Measurement Technology Conference (I2MTC). Piscataway:IEEE Press, 2020: 1-6.
[17]	TARIQ S , LEE S , SHIN Y ,et al. Detecting anomalies in space using multivariate convolutional LSTM with mixtures of probabilistic PCA[C]// Proceedings of Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery＆ Data Mining. New York:ACM, 2019: 2123-2133.
[18]	张克明, 蔡远文, 任元 . 基于生成对抗网络的航天异常事件检测方法[J]. 北京航空航天大学学报, 2019,45(7): 1329-1336.
	ZHANG K M , CAI Y W , REN Y . Space anomaly events detection approach based on generative adversarial nets[J]. Journal of Beijing University of Aeronautics and Astronautics, 2019,45(7): 1329-1336.
[19]	董彦磊, 汪春霆, 周萌 . AHP-BP卫星通信网评估算法应用研究[J]. 无线电工程, 2016,46(8): 9-13.
	DONG Y L , WANG C T , ZHOU M . Research on application of AHP-BP evaluation algorithm for satellite communication network[J]. Radio Engineering, 2016,46(8): 9-13.
[20]	CHENG S C , GAO Y , CAO J ,et al. Lecture notes in electrical engineering[M]. Singapore: Springer Singapore, 2020.
[21]	裴忠民, 任林涛, 熊伟 ,等. 基于社团结构的天基信息网络结构设计构想与仿真[J]. 天地一体化信息网络, 2020,1(2): 81-86.
	PEI Z M , REN L T , XIONG W ,et al. Design conception and simulation of space information network structure based on community structure[J]. Space-Integrated-Ground Information Networks, 2020,1(2): 81-86.
[22]	严佳洁, 祖家琛, 胡谷雨 ,等. 面向异构通信卫星的地球站组网规划方法[J]. 计算机科学, 2021,48(3): 275-280.
	YAN J J , ZU J C , HU G Y ,et al. Earth station network planning method for heterogeneous communication satellites[J]. Computer Science, 2021,48(3): 275-280.

领域	问题	问题难点	问题目标	原有方法	原有方法不足	智能方法
资源管理	资源分配	复杂动态的资源状况和业务需求	提高资源利用率，提高网络性能指标	建模并优化求解	复杂系统难以精确建模	DQN^{[3, 5-8]}，分布式强化学习^[4]
业务流量管理	流量预测	卫星互联网流量突发性、不稳定性	提高预测准确性	基于统计的时序模型	输入为单一维度，无法感知高维的网络上下文特征信息	LSTM^[10]，GRU^[11]，GCN-GRU^[12]
业务流量管理	流量调度	卫星互联网规模大、节点多、链路波动大	避免拥塞，降低平均时延	基于经验归纳调度规则	大规模网络使运维人员难以归纳调度规则	DQN^[13]，分布式强化学习^[14]
网络状态管理	异常检测	各类传感器数据维度高，异常可能在多点发生，周期性波动指标异常难以检测	提高检测准确率，降低虚警率、漏警率	基于经验设定正常范围	难以发现周期性波动指标异常，对指标逐一设定正常区间工作量大	聚类^[15]，LSTM^[16-17]，GAN^[18]
	网络评估	高维指标的选择、权重的确定	客观准确，减轻前期指标分析工作	基于经验确定指标权重	依赖专家经验，难以保障客观性，且权重无法调整	BP神经网络^[19-20]
	组网规划	大规模地球站和卫星平台节点	提高网络资源利用率	基于经验组网	资源利用率低	社团发现^[21-22]

方法名称	优势	适合的卫星互联网运维与管理问题
卷积神经网络	感知二维空间数据	资源分配等需考虑用户空间分布的问题^{[3, 5, 17]}
循环神经网络	感知时序数据	预测问题，如用户流量预测^[10-12]、信道质量预测、异常预测^[16-17]
图神经网络	感知图关系	学习卫星间拓扑关系^[12]，用户间、卫星与用户间的图关系
支持向量机	模型复杂度低	对计算效率要求高的场景，如核心网数据流异常检测
自编码器	高维数据降维	网络运维数据的高维特征降维；网络运行状态异常检测
生成式神经网络	生成模型，可应对小样本数据问题	解决数据样本不足或系统启动阶段尚未积累足够数据难点，也可以利用GAN生成同分布的数据，或用鉴别器检测异常^[18]
Q学习深度Q网络	解决无模型环境中的序列决策问题	动态网络环境中序列决策问题，如资源分配^[3-8]、流量调度^[13-14]
多智能体强化学习	各智能体独立决策	大规模网络中节点分布式控制资源分配^[4,14]、流量调度
迁移学习	所需训练样本少	网络有标签训练数据不足的问题