Description of the thesis topic

For ensuring low latency, Cloud resources are increasingly pushed to the edge of communication networks, bringing them closer to connected objects and users than to the usual Cloud data centers. These emerging Edge or Fog Computing architectures are characterized by their very large scale, their dynamic nature and the heterogeneity of their resources. As fog computing represents a major shift towards massive decentralization of Internet infrastructures, this trend is accompanied by an increase in the number and heterogeneity of computing and storage resources, which in turn increases the complexity of effective infrastructure management. In addition, the geographical dispersion of Fog resources means that computing capacity is limited locally, adding another dimension to the already complex problem of optimizing the overall energy consumption of the infrastructure. Indeed, this growing and worrying energy consumption is exacerbating the environmental impact of Fog infrastructures. Cloud providers are currently attempting to reduce this impact mainly by using energy from renewable sources, usually via photovoltaic panels managed by the Cloud providers themselves. However, renewable sources often offer intermittent and variable production at geographically dispersed sites. The aim of this thesis is to optimize the environmental impact of a fully distributed fog (without a centralized data center) powered by intermittent renewable energy sources. We will focus on the environmental impact of this Fog over its entire life cycle (manufacturing, transport, use and end-of-life). Possible applications for such a fully distributed, intermittently powered infrastructure will be explored.

Work Context

This PhD will be funded by CNRS on the CARECloud project and hosted in the Magellan team of the IRISA lab in Rennes. The PhD will be supervised by Anne-Cécile Orgerie (CNRS, IRISA).

Constraints and risks

The required skills are :
- a Master's degree in computer science or equivalent
- sound knowledge of optimization techniques and distributed systems or telecommunication networks
- knowledge of modeling and simulation
- knowledge/interest in digital sobriety
- a strong taste for research
- programming skills
- excellent communication and teamwork skills
- ability to write and report