Mir Muhammad Suleman Sarwar;Afaq Muhammad;Wang-Cheol Song
{"title":"IBN@Cloud:基于意图的云和叠加网络协调系统","authors":"Mir Muhammad Suleman Sarwar;Afaq Muhammad;Wang-Cheol Song","doi":"10.23919/JCN.2023.000051","DOIUrl":null,"url":null,"abstract":"This paper presents an intent-based networking (IBN) system for the orchestration of OpenStack-based clouds and overlay networks between multiple clouds. Clouds need to communicate with other clouds for various reasons such as reducing latency and overcoming single points of failure. An overlay network provides connectivity between multiple Clouds for communication. Moreover, there can be several paths of communication between a source and a destination cloud in the overlay network. A machine learning model can be used to proactively select the best path for efficient network performance. Communication between the source and destination can then be established over the selected path. Communication in such type of a scenario requires complex networking configurations. IBN provides a closed-loop and Intelligent system for cloud to cloud communication. To this end, IBN abstracts complex networking and cloud configurations by receiving an intent from a user, translating the intent, generating complex configurations for the intent, and deploying the configurations, thereby assuring the intent. Therefore, the IBN that is presented here has three major features: (1) It can deploy an OpenStack cloud at a target machine, (2) it can deploy GENEVE tunnels between different clouds that form an overlay network, and (3) it can then leverage the advantages of machine learning to find the best path for communication between any two clouds. As machine learning is an essential component of the intelligent IBN system, two linear and three non-linear models were tested. RNN, LSTM, and GRU models were employed for non-linear modeling. Linear regression and SVR models were employed for linear modeling. Overall all the non-linear models outperformed the linear model with an 81% R\n<sup>2</sup>\n score, exhibiting similar performance. Linear models also showed similar performance but with lower accuracy. The testbed contains an overlay network of 11 GENEVE tunnels between 7 OpenStack-based clouds deployed in Malaysia, Korea, Pakistan, and Cambodia at TEIN.","PeriodicalId":54864,"journal":{"name":"Journal of Communications and Networks","volume":"26 1","pages":"131-146"},"PeriodicalIF":2.9000,"publicationDate":"2024-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10416318","citationCount":"0","resultStr":"{\"title\":\"IBN@Cloud: An intent-based cloud and overlay network orchestration system\",\"authors\":\"Mir Muhammad Suleman Sarwar;Afaq Muhammad;Wang-Cheol Song\",\"doi\":\"10.23919/JCN.2023.000051\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper presents an intent-based networking (IBN) system for the orchestration of OpenStack-based clouds and overlay networks between multiple clouds. Clouds need to communicate with other clouds for various reasons such as reducing latency and overcoming single points of failure. An overlay network provides connectivity between multiple Clouds for communication. Moreover, there can be several paths of communication between a source and a destination cloud in the overlay network. A machine learning model can be used to proactively select the best path for efficient network performance. Communication between the source and destination can then be established over the selected path. Communication in such type of a scenario requires complex networking configurations. IBN provides a closed-loop and Intelligent system for cloud to cloud communication. To this end, IBN abstracts complex networking and cloud configurations by receiving an intent from a user, translating the intent, generating complex configurations for the intent, and deploying the configurations, thereby assuring the intent. Therefore, the IBN that is presented here has three major features: (1) It can deploy an OpenStack cloud at a target machine, (2) it can deploy GENEVE tunnels between different clouds that form an overlay network, and (3) it can then leverage the advantages of machine learning to find the best path for communication between any two clouds. As machine learning is an essential component of the intelligent IBN system, two linear and three non-linear models were tested. RNN, LSTM, and GRU models were employed for non-linear modeling. Linear regression and SVR models were employed for linear modeling. Overall all the non-linear models outperformed the linear model with an 81% R\\n<sup>2</sup>\\n score, exhibiting similar performance. Linear models also showed similar performance but with lower accuracy. The testbed contains an overlay network of 11 GENEVE tunnels between 7 OpenStack-based clouds deployed in Malaysia, Korea, Pakistan, and Cambodia at TEIN.\",\"PeriodicalId\":54864,\"journal\":{\"name\":\"Journal of Communications and Networks\",\"volume\":\"26 1\",\"pages\":\"131-146\"},\"PeriodicalIF\":2.9000,\"publicationDate\":\"2024-01-29\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10416318\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Communications and Networks\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/10416318/\",\"RegionNum\":3,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"COMPUTER SCIENCE, INFORMATION SYSTEMS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Communications and Networks","FirstCategoryId":"94","ListUrlMain":"https://ieeexplore.ieee.org/document/10416318/","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
IBN@Cloud: An intent-based cloud and overlay network orchestration system
This paper presents an intent-based networking (IBN) system for the orchestration of OpenStack-based clouds and overlay networks between multiple clouds. Clouds need to communicate with other clouds for various reasons such as reducing latency and overcoming single points of failure. An overlay network provides connectivity between multiple Clouds for communication. Moreover, there can be several paths of communication between a source and a destination cloud in the overlay network. A machine learning model can be used to proactively select the best path for efficient network performance. Communication between the source and destination can then be established over the selected path. Communication in such type of a scenario requires complex networking configurations. IBN provides a closed-loop and Intelligent system for cloud to cloud communication. To this end, IBN abstracts complex networking and cloud configurations by receiving an intent from a user, translating the intent, generating complex configurations for the intent, and deploying the configurations, thereby assuring the intent. Therefore, the IBN that is presented here has three major features: (1) It can deploy an OpenStack cloud at a target machine, (2) it can deploy GENEVE tunnels between different clouds that form an overlay network, and (3) it can then leverage the advantages of machine learning to find the best path for communication between any two clouds. As machine learning is an essential component of the intelligent IBN system, two linear and three non-linear models were tested. RNN, LSTM, and GRU models were employed for non-linear modeling. Linear regression and SVR models were employed for linear modeling. Overall all the non-linear models outperformed the linear model with an 81% R
2
score, exhibiting similar performance. Linear models also showed similar performance but with lower accuracy. The testbed contains an overlay network of 11 GENEVE tunnels between 7 OpenStack-based clouds deployed in Malaysia, Korea, Pakistan, and Cambodia at TEIN.
期刊介绍:
The JOURNAL OF COMMUNICATIONS AND NETWORKS is published six times per year, and is committed to publishing high-quality papers that advance the state-of-the-art and practical applications of communications and information networks. Theoretical research contributions presenting new techniques, concepts, or analyses, applied contributions reporting on experiences and experiments, and tutorial expositions of permanent reference value are welcome. The subjects covered by this journal include all topics in communication theory and techniques, communication systems, and information networks. COMMUNICATION THEORY AND SYSTEMS WIRELESS COMMUNICATIONS NETWORKS AND SERVICES.