LSTM-assisted Traffic Forecasting for Path Selection in an Overlay Network between Multiple Clouds at TEIN

다국어 초록 (Multilingual Abstract)

This paper presents an intelligent traffic management mechanism for an overlay network between OpenStack-based multiple Clouds at Trans-Eurasia Information Networking (TEIN). This paper shows three major contributions i.e., deployment of an overlay network between OpenStack-based Clouds at TEIN, deployment of a monitoring system for data set generation, and intelligent traffic management by RYU SDN-Controller using the output of the machine learning model. In the overlay network, there can be multiple paths of communication between a source and a destination node. Machine learning models can find the best path from a source to a destination. The data obtained from monitoring tools is used for model training. We utilize Long short-term memory (LSTM) and Linear Regression for traffic prediction of each link in the topology. Overall LSTM shows better performance with an 81% score as compared to the Linear Regression model which shows a 56% score. SDN-Controller ensures communication over the best path suggested by the ML model.

참고문헌 (Reference)

1 "perfSONAR"

2 A. M. Mir Muhammad Suleman Sarwar, "Vxlan tunneling in openstack based multi-site cloud a secure mechanism for communication and data sharing" 76-78, 2022

3 M. D. Mahalingam, "Virtual extensible local area network (VXLAN): A framework for overlaying virtualized layer 2 networks over layer 3 networks"

4 T. Saad, "Tunneling techniques for end-to-end VPNs: generic deployment in an optical testbed environment" 44 (44): 124-132, 2006

5 OpenStack, "The Most Widely Deployed Open Source Cloud Software in the World"

6 "Ryu SDN Controller"

7 Z. Aqun, "Research on tunneling techniques in virtual private networks"

8 R. G. Fielding, "RFC2616: Hypertext Transfer Protocol--HTTP/1.1"

9 Pypi, "Python api client"

10 "PySpark Documentation"

11 "Prometheus Pushgateway"

12 "Openstack"

13 M. Kimmerlin, "Network expansion in OpenStack cloud federations"

14 R. Chayapathi, "Network Functions Virtualization (NFV) with a Touch of SDN: Netw Fun Vir" Addison-Wesley Professional 2016

15 D. Boru, "Models for efficient data replication in cloud computing datacenters" 6056-6061, 2015

16 Canonical, "Microstack"

17 X. Su, "Linear regression" 4 (4): 275-294, 2012

18 염성웅 ; 김형태 ; 콜레카르 산자이 시바니 ; 김경백, "LSTM 기반 멀티스텝 트래픽 예측 기법 평가" 한국통신학회 24 (24): 13-23, 2021

19 미르 무함마드 술래만 사르워 ; 송왕철 ; AFAQ MUHAMMAD ; Sajid Alam, "IBN@TEIN: An AI-driven Intent-based Networking Platform for Service Deployment with QoS Assurance" 한국통신학회 25 (25): 1-9, 2022

20 Talha Ahmed Khan ; AFAQ MUHAMMAD ; 기자르 아바쓰 ; 송왕철, "IBN-based: AI-driven Multi-Domain e2e Network Orchestration Approach" 한국통신학회 23 (23): 29-41, 2020

21 J. Gross, "Geneve: Generic network virtualization encapsulation"

22 GeeksforGeeks, "Dijikstra"

23 K. J. Subramanian, "Data Security in Single and Multi Cloud Storage–An Overview" 4 : 19046-19052, 2016

24 P. Wang, "Cost-effective and latency-minimized data placement strategy for spatial crowdsourcing in multi-cloud environment" 11 (11): 868-878, 2023

25 "Asi@Connect"

26 R. Vinayakumar, "Applying deep learning approaches for network traffic prediction"

27 S. Shakya, "An efficient security framework for data migration in a cloud computing environment" 1 (1): 45-53, 2019

28 CISCO, "All Tunnels Lead to GENEVE"

29 N. Mansouri, "A new prefetching-aware data replication to decrease access latency in cloud environment" 144 : 197-215, 2018

30 S. Wang, "A network traffic prediction method based on LSTM" 17 (17): 19-25, 2019

31 S. Gopinath, "A comprehensive survey on data replication techniques in cloud storage systems" 13 (13): 15926-15932, 2018

32 A. Panarello, "A Requirements Analysis for IaaS Cloud Federation"