Implementation of Machine Learning for Autonomic Capabilities in Self-Organizing Heterogeneous Networks

The 3GPP’s self-organizing networks (SONs) standards are a huge step towards the autonomic networking concept. They are the response to the increasing complexity and size of the mobile networks. This paper proposes a novel scheme for SONs. This scheme is based on machine learning techniques and additionally adopting the concept of abstraction and modularity. The implementation of these concepts in a machine learning scheme allows the usage of independent vendor and technology algorithms and reusability of the proposed approach for different optimization tasks in a network. The scheme is tested for solving an energy saving optimization problem in a heterogeneous network. The results from simulation experiments show that such an approach could be an appropriate solution for developing a full self-managing future network.

• Plamen Semov
• Hussein Al-Shatri
• Krasimir Tonchev
• Vladimir Poulkov
• Anja Klein

Wireless Personal Communications, 2016.

https://link.springer.com/article/10.1007/s11277-016-3843-2