Evaluating the use of simultaneous cellular connections for reliable emergency communication

This project studies communication traffic across two concurrent cellular connections to quantify the potential gains of using two connections to increase reliability for emergency communications.
Master

This project has two parts. An experimental measurement part and an analysis part.

The measurement part involves using the MONROE Alliance test infrastructure to run experiments sending emergency communication modelled traffic simultaneously across two cellular connections, measuring what happens to each packet and recording the radio dynamics of the cellular connections. (See for more details). The experiments will use both static and moving nodes.

The analysis part of this project will analyse the data collected in the experiments to find what influences communication quality on each connection and how much combining both connections can improve quality. It will involve answer questions such as:

 

  • How often does the quality drop below acceptable levels?
  • How often do both channels drop in quality at the same time?
  • Can Machine Learning techniques be used to (a) better predict drops in quality (based e.g. on past performance, on radio metadata...), (b) steer packets onto one connection or the other?

Goal

To quantify the potential gains of using two connections to increase reliability for emergency communications.

Learning outcome

  • Learn about running experiments on real communication networks
  • Learn how to analyse data to discover interesting and correlated features.

Qualifications

  • Knowledge of networking is required.
  • Setting up experiments will require python coding and Linux skills.
  • Analysing the data will require skills in python or matlab or R or Julia, however skills in python coding are enough.

Supervisors

  • David Hayes
  • David Ros
  • Özgü Alay

 

Contact person