Space-time predictions of telecom traffic based on Apache Spark
The telecommunication industry is facing great challenges in modern times since more and more services are reliant on telecom infrastructure. To meet the new challenges and stay concurrent on the market it needs to evolve and start exploiting new technologies which is sometimes difficult since telecom industry is based on large traditional information systems. Also, since telecom is working with sensitive user data, they need to follow strict data sharing procedures to preserve the highest level of privacy for the users.
Having all that in mind, we developed Machine Learning pipeline based on Apache Spark that uses anonymized and aggregated data to predict day-time patterns in telecom data and to predict dynamic of telecom interactions based on spatial context. Our pipeline is developed using Apache Spark platform and programming languages Scala and Python to meet high performance demands.
With our solution, telecom provider is able to accurately predict traffic overload in their network, to identify “hot spots” in the urban areas where specific demands are present and to optimize services and infrastructure usage based on analytics.
If you would like to learn more about our projects, contact us at info@navanait.com
or fill the contact form.
