Advanced methods for traffic control in data centers
keywords CLOUD COMPUTING, CONGESTION CONTROL, PROGRAMMABLE DATA PLANES, P4, SOFTWARE DEFINED NETWORKING
Reference persons ANDREA BIANCO, PAOLO GIACCONE
Research Groups Telecommunication Networks Group
Thesis type SIMULATION AND DESIGN
Description Data centers offer a unique networking environment for cloud operators to improve the control of the traffic with respect to the legacy transport protocols adopted in the Internet. In order to minimize the delays experienced by the traffic, many advanced control schemes have been recently proposed, experimented and operated in large data centers.
The thesis will focus on the design and simulation of novel scheduling algorithms, queueing policies and traffic control systems that allows to improve the performance with respect state-of-art solutions. The solutions will be also validated in an emulated environment exploiting programmable data planes based on P4 programming model (https://p4.org/), which represents nowadays the most advanced technology to support Software Defined Networking (SDN).
Required skills Very strong motivation.
Academic background on computer networking.
Experience in programming.
Deadline 21/03/2021 PROPONI LA TUA CANDIDATURA