Software defined networking routing with machine learning

• Grant number: 19/03268-0
• Support Opportunities: Scholarships in Brazil - Master
• Start date: July 01, 2019
• End date: February 29, 2020
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Nelson Luis Saldanha da Fonseca
• Grantee: Daniela Maria Casas Velasco
• Host Institution: Instituto de Computação (IC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 15/24494-8 - Communications and processing of big data in cloud and fog computing, AP.TEM

Abstract

In the networking area, efficient traffic control such as routing appears as a critical challenge. Such challenge remains present since the conventional routing protocols (e.g., OSPF and RIP) currently present limitations. For instance, such protocols cannot adapt their decisions after degradation scenarios such as congestion, they are not intelligent. In this sense, protocols repeatedly continue to make the same routing decisions for similar congestion scenarios. Therefore, an intelligent network traffic control method is essential to meet this challenge. SDN decouples network control from packet forwarding, which significantly simplifies the switches operation. The good programmability of SDN also improves network feasibility and allows providing intelligence within the networks, making it easier to apply ML techniques. However, although some ML methods have been widely applied to network problems, such as traffic classification and traffic forecasting, ML remains presenting limitations. For instance, one of these limitations lies in defining which data can be collected and which control actions can be performed on network devices. Therefore, the ability to program the network using SDN and the advantages of ML that allow improving the understanding and performance of a system, alleviate these limitations and can be used to assist in the automation of network tasks such as routing. This work intends to take advantage of ML techniques and SDN to overcome the limitations of traditional routing protocols in networking. Thus, this proposal aims at developing an intelligent routing mechanism that can learn from previous experiences and adapt itself to the changes in SDN. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

CASAS-VELASCO, DANIELA M.; RENDON, OSCAR MAURICIO CAICEDO; DA FONSECA, NELSON L. S.. DRSIR: A Deep Reinforcement Learning Approach for Routing in Software-Defined Networking. IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT, v. 19, n. 4, p. 14-pg., 2022-12-01. (19/03268-0, 15/24494-8)

VILLOTA-JACOME, WILLIAM F.; RENDON, OSCAR MAURICIO CAICEDO; DA FONSECA, NELSON L. S.. Admission Control for 5G Core Network Slicing Based on Deep Reinforcement Learning. IEEE SYSTEMS JOURNAL, v. 16, n. 3, p. 12-pg., 2022-05-23. (19/03268-0, 15/24494-8)

CASAS-VELASCO, DANIELA M.; RENDON, OSCAR MAURICIO CAICEDO; DA FONSECA, NELSON L. S.. Intelligent Routing Based on Reinforcement Learning for Software-Defined Networking. IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT, v. 18, n. 1, p. 870-881, MAR 2021. (15/24494-8, 19/03268-0)

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

VELASCO, Daniela Maria Casas. Roteamento baseado em aprendizagem por reforço para redes definidas por software. 2020. Master's Dissertation - Universidade Estadual de Campinas (UNICAMP). Instituto de Computação Campinas, SP.