There is much hype in telecom. In just about every aspect of the network, new trends are shaping up: from the core to the radio access and through the transport network; in hardware, in software and in processes, buzz words and acronyms are aplenty. Worse, how often these acronyms are used out of context just to latch on a popular wave in a vain hope of marketing advantage? No wonder one can feel confused, even discouraged at times. What is real and what’s not? Who can really tell when there’s so much noise that masks real progress!
With this background, I want to stop in this post on data sciences, analytics and Big Data. These are fields that have slowly quickly built their way into the telecom industry as they did in other industries. McKinsey calls Big Data the next frontier for innovation, competition and productivity. A walk down the aisles in Mobile World Congress gives you a clear view of how big data analytics is with so many companies jumping on the Big Data bandwagon irrespective of whether they truly employ Big Data framework or a traditional data management framework. This year’s show saw major presence of major players in the analytics and Big Data field such as IBM, Oracle and HP which all combine strong computing platforms in their product portfolio. The wireless industry does not stand insular of other industries, but it is increasingly becoming a coupled with the compute and data processing sectors.
This should come to no surprise given the wealth of information mobile network generates on all levels: the network operation or the subscriber behavior and market levels. Today, it is thought that the information operators possess about their subscribers is key to thwarting the threat of over-the-top service providers. Similarly, information on network operation is key to potential cost savings and efficiency in operational processes as well as key to enhancing performance. Making sense of large amounts of unstructured and dynamic data from different sources that does not fit traditional models is big business.
The applications of data sciences in telecom are wide: from characterizing and detecting fraud and optimizing pricing to optimizing the performance of the wireless network. An example, consider NSN’s recent announcement on predictive operations where self-learning analysis engine evaluates data from different sources including service quality, customer experience and key performance indicators from the operator’s network, as well as external data such as weather forecasts and social media to detect abnormal patters up to 48 hours in advance with up to 95% accuracy and alert operation staff of the anomaly to take preemptive action. NSN is not alone in looking at how data sciences and Big Data can give it an edge in a market where hardware is a commodity that’s purchased from the lowest cost supplier. Information and intelligence is where differentiation lies that can command additional margins.
Of all the buzzwords in telecom, data sciences and Big Data techniques are for real. They represent the next step in the evolution of different tools ranging from performance management to customer assurance. The promise for operators is greater insight into the network. The promise for vendors is solidification of their position through high-value added features and services. This is an area that we at Xona Partners have focused on by combining telecom knowledge with data science expertise gained through years of developing machine learning algorithms and rolling out cloud management platforms, that have been applied in different fields including the Internet sector as well as the financial, medical and other sectors. To find out more on data sciences and how it fits the mobile operator framework, refer to our white papers Data Sciences Focus: Mobile Ecosystem Contributions, Data Sciences – Financial Sector Focus, and IT Transformation Towards a Big Data Environment.