@inproceedings{3e945358743d40c9a779e971ec108236,
title = "Performance Analysis of Multipath Deployment in Software-Defined Wide Area Network (SDWAN)",
abstract = "The high demand for network availability and reliability including wide-Area networking (WAN) brings many enterprises to use multiple links on their network. Due to the high cost of multiprotocol label switching (MPLS), many enterprises use a combination between MPLS and tunnels over the internet to build WAN infrastructure. With multiple links, critical applications required in business processes at the branch office can run properly on the occurrence of disruption in the primary link. In traditional WAN architecture, there is a routing mechanism to select the best path from the available link as common dynamic routings such as Open Shortest Path First (OSPF) and Border Gateway Routing Protocols (BGP). This paper analyzes the performance between traditional WAN and Software-Defined Wide Area Network (SDWAN) architecture in throughput, delay, and packet loss from a branch to headquarter. We used Ryu controller as an SDN Controller and Mininet as a data plane device to simulate SDWAN architecture. The resulting experiment showed that SDWAN architecture with the multipath link would have higher throughput than using a single link. For 10 Mbps bandwidth in each link, the total throughput is almost 20 Mbps, instead of 10 Mbps in a single link. An additional link monitoring application on the top SDN controller increased the performance in terms of jitter and packet loss in branch when the congestion occurred.",
keywords = "Mininet, multipath, Ryu controller, SDN, SDWAN, WAN, wide-Area networking",
author = "Rakhmat Yuniarto and Sari, {Riri Fitri}",
note = "Funding Information: ACKNOWLEDGMENT We thank the Ministry of Research and Technology research for financial support for this research under the PTUPT Grant number NKB-289/UN2.RST/HKP.05.00/2021. Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 International Conference on Artificial Intelligence and Computer Science Technology, ICAICST 2021 ; Conference date: 29-06-2021",
year = "2021",
month = jun,
day = "29",
doi = "10.1109/ICAICST53116.2021.9497800",
language = "English",
series = "ICAICST 2021 - 2021 International Conference on Artificial Intelligence and Computer Science Technology",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "124--128",
booktitle = "ICAICST 2021 - 2021 International Conference on Artificial Intelligence and Computer Science Technology",
address = "United States",
}