Experiment Scheduling in Large Scale Platforms: Fairness, Complexity and Efficiency


MONROE project has developed a unique distributed platform to conduct independent, repeatable, multi-homed, large-scale measurement and experimental campaigns for collecting data from operational Mobile Broadband (3G and 4G) networks.

The platform has many users with different experiment requirements. To serve its experimenters, MONROE designed an experiment scheduling system that ensures that there are no conflicts between experimenters when running their experiments and, assigns a time-slot and equal resources to each experimenter. The scheduling system is designed to be robust against short-time loss of connectivity .


The goal of this project is to design and implement algorithms that will ensure the fairness to the experimenters while taking complexity and efficiency into account. This indeed is an optimization problem: given the available resources (data quotas and time) as well as experimenters need (data, CPU, processing power and duration), what is a fair distribution of resources to experimenters? The project will investigate different methodologies considering complexity and efficiency.

Learning outcome

  • Different task scheduling methodologies
  • Hands on experienced in real-world systems
  • Familiarity with optimization frameworks


  • Strong analytical skills
  • Programming skills (Python)
  • Experience with Linux


Özgü Alay

Collaboration partners

Celerway Communication AS