Monetizing Telecom Big Data - A Strategic Approach for the Operators

Introduction

In search of new revenue streams, harnessing the potential of Big data seems to be on the agenda of every mobile operator. Terabytes of data lying with the operators remain unutilized and this rich volume of data accessible by the mobile operators provides an immense opportunity in monetization.

Today, the operators are looking out for a better usage with the unutilized data, that’s to sort, analyze and present the data for monetization. Unavailability of a proper system to orchestrate all these structured and unstructured data from various sources for monetization has been a great challenge. This article provides an insight to the operators and vendors for effective monetization using the widely used Open source Hadoop platform

Opting for a Cloud based solution instead of an On-Premise Big data infrastructure helps the operators to avoid CAPEX investments and switch to an OPEX Model. Monetization and presentation of relevant data can be done based on a predefined data catalog. Amazon’s Elastic Map Reduce service can be used for Big Data processing.

Data Collection

The Apache Flume can be used to collect  all important signaling information to facilitate monetization.The flume Agents after collecting streaming data can be send to the  Designated HDFS storage. HDFS   (Hadoop Distributed File System) a Hadoop cluster, here, acts as the distributed storage house that can store the data be it from the EPC Source using Big Data Client Application or from the EPS Data Source using the Apache Open Source Flume framework

Apache Flume, a distributed data collection service aggregates and moves the large amounts of streaming data into HDFS. The data so being collected at the HDFS are then processed using either the Apache Pig or Apache Hive which acts as a data warehouse providing query, analysis and data summarization. The extracted data can then be used for analytical applications. The entire system can be implemented on a Cloud based infrastructure.

Flume Deployment Schematic

Here the collector is also known as Data Sink. 

The traffic flowing through the 4G control plane contains the most valuable and strategic information in the network such as – the location of the subscribers, their phone number, the kind of handset they use to connect to the network, available bandwidth, charging methodology, IP Address, services used, the remaining balance and so on.  

Applications of Telecom Big Data

Big Data find its application in various domains.

Fraud Management is one such area where Big data plays a vital role, with the real time call data analysis to identify fraud immediately and take proactive actions based on the same. Big data also fits its role in optimizing routing and quality of service by analyzing network traffic in real time. Apart from this, Big data also plays a notable role in Mobile advertising, M2M, Business Intelligence, Legal Intercept and so on.

How to create Additional Revenue Streams using Telecom Big Data?

Operators can rent out the data to a third party application and Business entities thereby potentially creating additional revenue opportunities for it's usage.  The information flowing through the 4G Signaling control plane can be categorized into three Major areas:-

• Subscriber Authentication information
• Policy Information
• Billing Information

 Some of the important fields are mentioned below :-

The Operators need not invest heavily on acquiring new custom made hardware’s to deploy the Big data platform. The Elastic Map Reduce service of Amazon Cloud can be used for the Big Data Processing and Analytic's. 

Apache Hadoop is an open-source system that assures reliable storage and processing of information across many commodity computers. It consists of two core components,

• HDFS : For Distributed and Clustered Storage
• Map/Reduce: Fault-tolerant distributed processing.

The Hadoop platform can be a deployed on an array of inexpensive commodity Hardware’s. The Data Replication and Hardware Redundancies are built as part of the Hadoop platform. If the Operator is willing to look at the Elastic Map Reduce service of Amazon Cloud, then the heavy lifting related with setting up the Hadoop Cluster  can be avoided. 

Background to this Paper

Original Research, Reference and Acknowledgement

This article was written based on an independent research to bring out the monetization opportunities available in the Network- I could only touch the tip of the iceberg in this article. Would like to acknowledge the efforts of Arun Kumar and Pinaki Mukherjee in collecting and refining few data points for me.

https://in.linkedin.com/pub/pinaki-mukherjee/8/681/b08

 Your comments and opinions are always welcome.  Please feel free to share this article to the folks who are working on Big Data, Cloud, Telecom, or Mobile Advertising space. The ideas shared here can be applied across various domains.