Cloud-Based Implementation of an Automatic Coverage Estimation Methodology for Self-Organising Network

Daniel Fernandes, Diogo Clemente, Gabriela Soares, Pedro Sebastião, Francisco Cercas, Rui Dinis, Lucio S. Ferreira

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)
2 Downloads (Pure)


One of the main concerns of telecommunications operators is related to network coverage. A weak coverage can lead to a performance decrease, not only in the user experience, when using the operators' services, such as multimedia streaming, but also in the overall Quality of Service. This paper presents a novel cloud-based framework of a semi-empirical propagation model that estimates the coverage in a precise way. The novelty of this model is that it is automatically calibrated by using drive test measurements, terrain morphology, buildings in the area, configurations of the network itself and key performance indicators, automatically extracted from the operator's network. Requirements and use cases are presented as motivations for this methodology. The results achieve an accuracy of about 5 dB, allowing operators to obtain accurate neighbour lists, optimise network planning and automate certain actions on the network by enabling the Self-Organising Network concept. The cloud implementation enables a fast and easy integration with other network management and monitoring tools, such as the Metric platform, optimising operators' resource usage recurring to elastic resources on-demand when needed. This implementation was integrated into the Metric platform, which is currently available to be used by several operators.

Original languageEnglish
Article number9060956
Pages (from-to)66456-66474
Number of pages19
JournalIEEE Access
Publication statusPublished - 2020


  • Cloud implementation
  • coverage estimation
  • drive tests
  • measurements
  • propagation model


Dive into the research topics of 'Cloud-Based Implementation of an Automatic Coverage Estimation Methodology for Self-Organising Network'. Together they form a unique fingerprint.

Cite this