Two Master Thesis opportunities at Ericsson Research

Skip to first unread message

Börje Ohlman

Nov 2, 2021, 6:40:22 AM11/2/21

If you can help spreading these, in my view, very interesting Master Thesis opportunities to interested students that would be great.

Anyone interested should in applying for the jobs should follow the procedure described in the below linked ads.




Master Thesis - Natural language system to generate Terraforms

Thesis Objectives:

Use a natural language understanding system to generate Terraforms* for configuring infrastructure of a data collection/management system.


There is ongoing work to use Google functions to build an infrastructure for a data collection/management system. The task of this thesis work is to build a natural language front end to this system using a natural language understanding system.


A Data Collection Architecture is used for collecting data from mobile communication systems. The data can be used for management and operations purposes as well as for AI/ML learning processes.


Master Thesis - Using object security to secure data

Thesis Objectives:

In many of today's telecom systems data is secured by being stored in secured servers that communicate via secured pipes using e.g. TLS. This system has several vulnerabilities, including

  • server penetration
  • those having access to a server can often access all the data on the server
  • issues when transferring data between trust domains
  • Man-in-the-middle attacks

 As an alternative, data can be secured as individual information objects (IO) by the publisher at the same time as the IO is published.  

Data that should be possible to secure includes, but are not limited to:

  • Data and measurement logs
  • Customer data
  • Management commands
  • Data for AI training and machine learning 

Possible techniques to use:

  • Google IAM
  • Attribute based encryption (ABE)

The thesis should develop/select/evaluate a method where data is secured by the publisher at the time of data creation, but is still easily accessible for legitimate data consumers.


Reply all
Reply to author
0 new messages