Self-Organizing Networks (SON) are currently being introduced by the 3rd Generation Partnership Project (3GPP) as part of the Long Term Evolution (LTE) system as a key driver for improving the operation of wireless networks. Given that many challenges have been identified when moving from the SON concept to practical implementation, this article claims the suitability to gain insight into the problem by taking advantage of optimization mechanisms in live Universal Mobile Telecommunications System (UMTS) networks. From that perspective, this article outlines a roadmap from actual manual optimization toward the inclusion of SON concepts in future networks. In particular, an optimization framework is developed considering different stages that include the collection of inputs from different sources (network counters, measurement reports, drive tests), the tuning parameters to achieve optimization, and the optimization procedure itself. Even though the proposed framework is presented and evaluated for convenience in a UMTS context, it is rather generic and technology-agnostic and, therefore, it can be formulated as an initial reference for LTE self-optimization use cases. The proposed framework is applied to a practical coverage optimization use case supported by data extracted from a real UMTS network in a European city, illustrating the capability to automatically identify a cell with sub-optimal coverage and to provide a solution to the problem. The extension of the case study to LTE is also analyzed. Finally, as a result of the lessons learned, the article makes a projection to the LTE context by identifying the key points to be solved for the materialization of a self-optimization procedure.

Peer Reviewed