ESR10

Cosmin Florin Avasalcai

• Current Position: Software/System Architect, Siemens
• Education:
  • Ph.D. Computer Science, TU Wien, Austria
  • M.Sc. Computer Science and Engineering,  Technical University of Denmark, Denmark
• Research Interests:
  • Edge Computing
  • Resource optimization
  • Cloud Computing

Research project status:

Ended in 2021

c.avasalcai@dsg.tuwien.ac.at

LinkedIn Profile

Quality of Service aware Resource Provisioning in Edge Computing

Host: Vienna University of Technology (TU Wien), Distributed Systems Group

• Main supervisor: Prof. Schahram Dustdar, dustdar@dsg.tuwien.ac.at (contact person)
• Co-supervisor: N.N -> formerly Assistant Prof. Stefan Schulte, s.schulte@infosys.tuwien.ac.at

Objectives:

1. Enable resource provisioning for large-scale Fog environments, considering their volatility.
2. Develop resource provisioning and management techniques for resource allocation under given Quality of Service (QoS) constraints targeting non-critical applications, ensuring non-interference with critical apps.

Expected Results:

• Models for Fog resources, and definition of optimization problems and metrics for resource provisioning.
• Development of prototype provisioning mechanisms for self-adaptation, supporting the fusion and dissolution of groups of Fog Nodes, data/state migration, sharing, replication, as well as the resource allocation services.
• Development of mechanisms for self-optimizing resource allocation and task scheduling.
• Evaluate the devised approaches, via simulations and real-world experiments.
• Develop services that enable pricing plans used in cost-to-serve infrastructure use.

Visits and Collaboration:

• MDH (Assoc. Prof. Moris Behnam): Define the API for the resource management jointly with MDH and ABB
• DTU (Prof. Paul Pop): Define a resource management framework to efficiently host IoT applications at the edge of the network.

Description:

Within this Ph.D. project, the candidate is expected to develop optimization approaches for resource allocation and task scheduling in the fog. For this, the candidate is expected to apply techniques from the fields of Machine Learning and/or (Linear) Optimization. The work conducted within this project will follow an iterative approach, starting with centralized, optimal solutions but later on working on decentralized, optimal, and heuristic approaches.

Recent publications:

For a complete list of publications please visit the following page: Publication Page