Nokia has signed a global agreement with Hewlett Packard Enterprise (HPE) to license technology assets that will add to its AI-driven RAN automation capabilities. The deal expands Nokia’s MantaRay Service Management and Orchestration (SMO) portfolio.
As part of the agreement, Nokia gains access to HPE’s RAN Intelligent Controller (RIC) technology, which will be integrated into MantaRay’s AI-powered SMO and network automation suite. In addition, HPE’s development team responsible for these assets has joined Nokia Mobile Networks, effective October 1, 2025.
“This licensing deal with HPE will further strengthen our proven MantaRay SMO portfolio by adding these assets and expertise,” said Tommi Uitto, President of Mobile Networks at Nokia in a press note. “Our customers worldwide will benefit from enhanced capabilities in AI-driven automation, orchestration and open ecosystems, helping them manage multi-vendor networks more efficiently and prepare for the transition from 5G to 6G.”
Nokia’s MantaRay SMO is a platform for AI-powered RAN automation and orchestration. It already meets TM Forum’s Autonomous Networks Level 4 benchmark, supports Open RAN compliance, and is compatible with the open R1 interface for rApps.
The move comes at a challenging time for the broader RAN market, which has shrunk dramatically in the past two years. According to Omdia, global RAN product sales fell from about $45 billion in 2022 to $35 billion in 2024. Within that, Open RAN has struggled to gain traction despite industry hype.
Earlier this year, open RAN vendor Mavenir decided to exit from radio unit manufacturing. In addition, Echostar decided to decommission its US Open RAN build, and VodafoneThree UK’s opted to award major 5G contracts to Ericsson and Nokia while ignoring Samsung, which had deployed Open RAN at around 70 trial sites.
Against this backdrop, Nokia’s integration of HPE’s RIC technology signals a pragmatic approach: bolstering automation and interoperability in multi-vendor environments while advancing toward 6G-ready networks, without betting solely on the Open RAN model.
