Improving the Connectivity of Mobile Networks Through the Use of Deep Learning for Planning Neighborhood Relations

Abstract

This work proposes a Deep Learning approach to help mobile operators plan neighborhood relationships within their network in the form of an application allowing to map sites, sectors and their cells, to calculate the neighbors of different cells based on deep learning algorithms and integrate new sectors or sites on the network. By leveraging complex machine learning techniques, the tool developed can identify patterns, trends and hidden relationships between different base stations in the network. This leads to more effective and efficient planning of neighbor relationships, thereby reducing interference and improving quality of service for subscribers. Additionally, the tool offers increased flexibility and adaptability. It is able to adjust to rapid network changes, such as adding new base stations or changing a network site's settings. This ability to adapt guarantees continuous and optimal planning of neighborhood relations, even in a dynamic and evolving environment.