Defence of dissertation in the field of Computer Science, M.Sc. (Tech.) Kasper Apajalahti

How to find the best context- and objective-specific algorithm to optimize a mobile network?
CS_defence photo by Matti Ahlgren

According to the dissertation "Contributions to Self-Organizing Networks and Network Measurement Data Management" in the future, mobile networks need to facilitate wireless communication in many verti-cal domains, such as automotive, manufacturing, energy, healthcare, and end users’ devices. The diverse set of use cases increases the complexity of the network man-agement and maintenance.

Along with the 4G, the Self-Organizing Networks (SON) paradigm has already been designed and utilized to automate network management use cases. SON functions optimize, for example, energy consumption, capacity, and coverage of mobile net-works. However, SON functions are designed to specific use cases and contexts. For example, a SON function may be developed to one network to provide an excellent signal quality cost-effectively in an urban area during rush hours. Such a tried and tested SON function would be useful to share with other use cases and situations having similar objectives and contexts.

This thesis is conducted by designing, implementing, and evaluating techniques that provide more flexibility to utilize SON functions across networks, use cases, and con-texts. Two user interface functionalities are designed to 1) discover SON functions and 2) explore the performance and functionality of an individual SON function. Also, semantic models are designed to link network- and SON-specific information across networks and systems. Due to semantic modelling, suitable SON functions are easier to discover across networks. Moreover, two statistical methods are developed to facili-tate SON function modelling and SON function-specific measurement data modelling.

Opponent: Professor Bernhard Bauer, University of Augsburg, Germany

Custos: Professor Eero Hyvönen, Aalto University School of Science, Department of Computer Science

Contact information: Kasper Apajalahti, Department of Computer Science, [email protected]

Electronic dissertation:

  • Published:
  • Updated:
URL copied!