Expansion of Congestion Latency Optimization Algorithms

Organization

Universidad de Palermo

Tipo

Academic Sector

Años

2020

Paises

Argentina

The project will expand the congestion algorithm for latency optimization developed by Universidad de Palermo so that it can be executed from both the side receiving and the side sending the data. The algorithm currently works from the side receiving the data and optimizes the reception window to reach available network capacity without causing packet flooding at the bottleneck location on a network path, thus improving the latency of all the connections that share the bottleneck. The expansion will seek to develop a version of the congestion algorithm that can be executed from the side sending the data. This will optimize the “outgoing” traffic of the servers originating the data. The algorithm will also be incorporated to a standard Linux distribution. The funds provided by FRIDA will be used to train and incorporate additional researchers to the team currently working at the research center of Universidad de Palermo’s School of Engineering so as to increase its capacity for exploring alternatives and developing specific solutions to meet the goals above.