The research firm said machine learning technologies will lead operators to profoundly change how they manage the telecoms business. 

 “Machine learning-based predictive analytics are applicable to all aspects of the telecoms business,” said Joe Hoffman, managing director and VP at ABI Research. “It is important that operators master and internalize these technologies and not rely solely on their vendors’ expertise.”

Hoffman said machine learning can deliver benefits across operators’ operations with financially oriented applications, including fraud mitigation and revenue assurance, which currently make the most compelling cases. Legacy analytics are rule-based solutions that cannot keep pace with the criminal element, but machine learning excels at spotting trending anomalies.

He said predictive machine learning applications for network performance optimization and real-time management will introduce more automation and efficient resource utilization. Even sales, marketing, and customer experience teams will benefit as machine learning helps to innovate and reengineer business processes.

Telecoms big data solutions include the commercial IT kit; the open source, Java-based Hadoop ecosystem, SQL/NoSQL data management, and orchestration platforms. Spending on this infrastructure will exceed $7 billion in 2021. But the biggest growth and most value comes from using predictive analytics to improve telecom business performance, with machine-learning-based predictive analytics to grow at nearly 50% CAGR and reach $12 billion through 2021.

Leading infrastructure vendors—Ericsson, Huawei, Nokia and ZTE—are delivering big data and machine learning solutions oriented toward network operations. Even Hadoop/NoSQL startups like Argyle Data, and chip vendors, led by Intel and Qualcomm, are delivering solutions pertinent to the telecom operator.

“With the rise of commercial cloud infrastructure and machine learning services, every mobile operator can be a big data company,” concludes Hoffman. “In just a few years, we will see the mobile networks of tomorrow manifest into giant, distributed supercomputers, with radios attached, continuously reengineered by machine learning.”

Telecomasia.net