MEC-AI: Edge computing enhanced by artificial intelligence

Project description

Edge computing emerges as an increasingly important way to process data from networked devices and sensor networks into information that applications can utilise. Over the recent years, data has been processed in clouds on a centralised basis, but edge computing brings computing servers closer to the edge of the network, for example, locally next to 5G base stations. This makes it possible to transfer data faster from devices and systems to the servers that process it. Low latency is especially important in the control of industrial processes and machines. Edge computing also enhances data security: for example, industrial facilities or hospitals no longer need to transfer data far into the cloud, but instead the data will remain within the own facility.

Researchers at the University of Oulu seek to provide companies with example solutions that make use of edge computing and artificial intelligence. Artificial intelligence can learn the variations in data volumes and user movements, for example, and help servers to prepare for them by distributing the computing load and downloading content in advance. This way, networks will not be blocked by large user and data volumes, but instead the applications will run smoothly.
 
The principal aim of the project is to help Finnish companies in utilising new generation technology. The test environment will be provided with a first-generation commercial Mobile Edge Computing (MEC) system during 2017, with computing servers directly connected to 5G base stations. MEC is a standardised edge computing definition, and computing servers that comply with it are developed by companies such as Nokia. The research team designs artificial intelligence methods and extensions to the MEC system in collaboration with companies. An example application will be set up for the test environment to provide companies with tangible information on the opportunities offered by edge computing and artificial intelligence.

Project coordinator

University of Oulu

Project results

Articles in journals

Articles in conference proceedings

  • J. Haavisto, M. Arif, L. Lovén, T. Leppänen, J. Riekki, J. (2019) Open-source RANs in practice: an over-the-air deployment for 5G MEC. In: European Conference on Networks and Communications (EuCNC2019).
  • J. Islam, E. Harjula, T. Kumar, P. Karhula & M. Ylianttila (2019), Docker Enabled Virtualized Nanoservices for Local IoT Edge Networks, 5th IEEE Conference on Standards for Communications & Networking (IEEE CSCN'19), Oct 28-30, Granada, Spain. (to appear)
  • T. Kumar, A. Braeken, V. Ramani, I. Ahmad, E. Harjula & M. Ylianttila (2019), SEC-BlockEdge: Security Threats in Blockchain-Edge based Industrial IoT Networks, 11th International Workshop on Resilient Networks Design and Modeling (RNDM), Oct 14-16, Nicosia, Cyprus. (to appear)

Partners

  • Center for Ubiquitous computing (UBICOMP): Prof. Jukka Riekki, Lauri Loven, Teemu Leppänen
  • CWC Networks and systems: Associate prof. Mika Ylianttila, Dr. Erkki Harjula
  • Applied and Computational Mathematics (ACM): Prof. Mikko Sillanpää: Leena Ruha

Research groups

  • Center for Ubiquitous Computing
  • CWC-NS Networks and Systems
  • Applied and Computational Mathematics

People

Mika Ylianttila

Associate professor (tenure track)
Mikko Sillanpää

Mikko J. Sillanpää

Professor of Statistics
Lauri Lovén

Lauri Lovén

Doctoral Researcher

Erkki Harjula

Project Manager / Postdoctoral researcher
Johirul Islam

Johirul Islam

Doctoral Research