Vodafone Ireland Leverages Cisco C-SON for Machine Learning to Predict 3G Traffic Flows

“Innovation distinguishes between a leader and a follower.”

Steve Jobs

Vodafone engineers are constantly trying to innovate within their network to improve their customers’ experiences. To that end Vodafone Ireland has teamed with Cisco to create the world’s first trial using machine learning algorithms in Cisco Centralized Self Optimizing Network (C-SON) to predict where 3G traffic will peak in the following hour.

According to McKinsey & Company, the advent of machine learning and other advanced-analytics techniques, combined with the ability to now digitize large numbers of operator processes, creates a new paradigm that allows for an unprecedented change in operator cost structure. Applying analytics to the vast customer data available to operators delivers new insights about needs and preferences and what may cause customers to leave. By digitizing processes, for the first-time operators can act on this information both cost effectively and at scale, tailoring products, services, and interactions to individual customers. These new technologies are just as valuable in managing network infrastructure, guiding investments in new capacity, and adjusting wireless networks automatically.

C-SON provides a centralized architecture where the optimization algorithms reside in the network management system or a central SON server that manages all edge radio nodes. It can orchestrate the behavior of radio network equipment across the entire network of multi-vendor and multi-technology environments.

By monitoring the network traffic trends, to predict the future network behavior based upon data processing and pattern recognition. By doing this the network can self-configure itself automatically to balance the traffic load amongst neighboring cell sites and improve the customer experience.

According to Santiago Tenorio, Head of Network Strategy & Architecture at Vodafone, “Vodafone customers could experience significant benefits from the use of machine learning in our networks. For instance, the network could identify if there is high traffic at a mobile cell site every Thursday at 8pm – perhaps generated by weekly concerts at a popular music venue – and automatically increase the cell’s capacity before people arrive, returning to normal after they go home. Customers would benefit from the uninterrupted ability to call, message or share videos and photos on social media throughout the night. Initial results confirmed an average 6% improvement in the mobile download speed and lower interference at the cell sites (the cause of dropped calls, problems connecting and higher device battery drain).”