Advanced search
Start date
Betweenand

DigiNet: towards digital-twin assisted networks

Abstract

Data Plane Programmability is redesigning the way we manage and operate data forwarding devices. Recent efforts led by the industry and academia have shown a growing adoption of intelligent and autonomous mechanisms to deal with an increasing number of problems. Despite all the progress made in the area, the application of Artificial Intelligence (AI) in network infrastructures is not that simple. One of the fundamental questions behind the adoption of AI in critical systems consists of determining whether the models have been designed and trained correctly and faithfully to reality. An emerging technology to verify and validate AI models consists of using Digital Twins (DT). DTs allow to create digital representations of physical entities in the network infrastructure. These digital representations are used as realistic prototypes to infer, for example, the correctness and performance of existing autonomous mechanisms or for simple future inferences. Despite reducing the risks associated with deploying and operating AI-based mechanisms in network infrastructures, operating DT-assisted network infrastructures requires substantial efforts in orchestrating resources to keep up with the requirements of physical and digital entities. In this project, we propose DigiNet, the first effort to implement, operate, and orchestrate Digital Twins-assisted network infrastructures. The project unfolds into three objectives: (i) to design a framework for allowing the correct description Digital Twin Networks models; (ii) to optimize the management and operation of Digital Twin Networks on programmable infrastructures in an analytical and experimental way; and (iii) to identify problems in the networking domain that can benefit from the use of Digital Twin Networks. The results to be obtained in this project have the potential to cause a lasting impact within its research area in the coming years. (AU)

Articles published in Agência FAPESP Newsletter about the research grant:
More itemsLess items
Articles published in other media outlets ( ):
More itemsLess items
VEICULO: TITULO (DATA)
VEICULO: TITULO (DATA)

Scientific publications (4)
(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)
VOGT, FRANCISCO GERMANO; RODRIGUEZ, FABRICIO; LUIZELLI, MARCELO CAGGIANI; ROTHENBERG, CHRISTIAN ESTEVE. Rethinking the In-band Network Telemetry: Towards Application and Server-Level Network Telemetry. PROCEEDINGS OF THE CONEXT STUDENT WORKSHOP 2024, CONEXT-SW 2024, v. N/A, p. 2-pg., . (23/00794-9, 21/06981-0)
VOGT, FRANCISCO GERMANO; RODRIGUEZ, FABRICIO; LUIZELLI, MARCELO CAGGIANI; ROTHENBERG, CHRISTIAN ESTEVE. Poster: Towards In-Network Resource Scaling of VNFs. PROCEEDINGS OF THE 20TH INTERNATIONAL CONFERENCE ON EMERGING NETWORKING EXPERIMENTS AND TECHNOLOGIES, CONEXT 2024, v. N/A, p. 2-pg., . (21/06981-0, 23/00794-9)
VOGT, FRANCISCO GERMANO; CESEN, FABRICIO EDUARDO RODRIGUEZ; DE CASTRO, ARIEL GOES; SINGH, SUNEET KUMAR; LUIZELLI, MARCELO CAGGIANI; ROTHENBERG, CHRISTIAN ESTEVE; ANTICHI, GIANNI. Video Streaming QoE Meets Programmable Data Planes: The Case of In-Network QoE for 360°VR. IEEE NETWORK, v. 39, n. 2, p. 8-pg., . (23/00794-9, 21/06981-0, 21/00199-8)
LUIZELLI, MARCELO C.; VOGT, FRANCISCO G.; SEVERO DE SOUZA, PAULO SILAS; LORENZON, ARTHUR F.; DA COSTA FILHO, ROBERTO I. T.; ROSSI, FABIO D.; CALHEIROS, RODRIGO N.; ROTHENBERG, CHRISTIAN ESTEVE. DigiNet: Scaling up Provisioning of Network Digital Twin. 2024 IEEE 10TH INTERNATIONAL CONFERENCE ON NETWORK SOFTWARIZATION, NETSOFT 2024, v. N/A, p. 9-pg., . (23/00794-9, 21/06981-0)