How telecom providers use AI to cut costs and create new revenue

Telecommunications providers are rapidly embracing artificial intelligence as a dual-purpose technology that simultaneously reduces operational costs and creates new revenue opportunities. Fujitsu Networks executive Rich Colter highlighted at Mobile World Congress that the most successful AI implementations in telecom extend beyond just radio network optimization to encompass the entire network infrastructure, including optical networks and operational workflows. This comprehensive approach allows communications service providers (CSPs) to remain competitive while extracting maximum value from their substantial network investments.

The big picture: CSPs need an end-to-end AI strategy that addresses both network optimization and new business opportunities rather than implementing isolated AI solutions.

• “AI is critical for the network,” explained Fujitsu Networks Business Head of Global Marketing Rich Colter. “There’s opportunities in how you monetize the network, and also how we optimize and transform the network.”
• Open networking principles provide the foundation for this comprehensive approach, extending from Open RAN initiatives to optical networks with open line systems and Open ROADM.

Three-pronged AI approach: Fujitsu categorizes AI implementation in radio networks into distinct strategies that serve different business goals.

• AI and RAN involves shared infrastructure where both AI workloads and RAN workloads operate on the same hardware, allowing operators to “both manage your infrastructure and monetize spare capacity.”
• AI for RAN focuses on enhancing network performance, with Fujitsu’s customers seeing significant performance improvements of 20% to 50% across various optimization use cases.
• AI on RAN leverages the radio network to support generative AI applications, ensuring providers can meet evolving customer needs.

Practical implementations: Fujitsu has already begun deploying these concepts with major industry partners.

• At Mobile World Congress, Fujitsu demonstrated with Arrcus, Eviden, Liberty Global, NVIDIA and Philips how advanced consumer applications can run on AI-based Open RAN networks alongside traditional network functions.
• The company is conducting verification testing with SoftBank and NVIDIA on use cases like uplink channel interpolation, where AI significantly boosts radio network performance.

Beyond technology: Successful AI implementation requires organizational and workflow changes throughout the network lifecycle.

• AIOps methodologies use AI to optimize networks “throughout its lifecycle,” from initial deployment to day-to-day operations and troubleshooting.
• One practical example involves using AI for accelerated root cause analysis, allowing network operations centers to quickly identify underlying problems when issues arise.
• Fujitsu’s AI-driven network modernization tools have demonstrated impressive efficiency gains, reducing manual effort by 80%, required task hours by 75%, and delivery time by 67%.

Why this matters: As telecommunications evolves, AI is transforming from an optional enhancement to a fundamental competitive necessity for service providers looking to drive efficiency and develop new revenue streams.