Nokia Media Release | February 20, 2024
press release
Nokia and Qualcomm collaborate on AI interoperability technology to improve wireless capacity and performance #MWC24
- Qualcomm and Nokia Bell Laboratories have overcome the critical AI interoperability challenge that prevents collaboration between multivendor artificial intelligence (AI) systems in devices and networks.
- A technique called sequential learning allows different AI models to be trained independently, allowing vendors to adjust to each other without having to share their differentiating AI models.
- Qualcomm and Nokia Bell Laboratories have prototyped this sequential training solution and will demonstrate it at MWC24 in Barcelona.
Espoo, Finland and San Diego, California – Qualcomm Technologies and Nokia today announced that they have developed prototype technology that can solve one of the fundamental challenges facing the use of AI in wireless systems: interoperability. . By enabling multi-vendor interoperability between different AI models used in networks and devices, future wireless systems will support the use of AI and reduce energy consumption while increasing capacity and You will be able to achieve superior performance.
Qualcomm Technologies and Nokia Bell Labs implemented this technology in an AI proof of concept using Nokia Bell Labs' test network and Snapdragon 5G modem RF system. The companies plan to demonstrate this proof of concept at Mobile World Congress 2024 in Barcelona.
AI and machine learning will have a major impact on the future of wireless communications, as AI models can optimize radio performance for specific environments. However, for AI to reach its full potential in wireless networks, the various AI systems utilized in networks and devices must coordinate their activities. Otherwise, these AI systems may serve different purposes.
Qualcomm Technologies and Nokia Bell Laboratories have adopted a new approach to AI interoperability using a technique called sequential learning. With sequential learning, the network and device share data about connections related to AI training, but do not need to share the actual AI model. Separate encoders and decoders then use that data to independently train network AI and device AI. These independent encoders and decoders effectively act as translators, allowing multiple widely disparate AI systems to share knowledge with each other. Through a series of trainings, these different AI models, regardless of vendor, strengthen each other. Additionally, sequential training allows wireless handsets or devices from multiple vendors, each with their own encoder implementation, to interact with the same base station with a common decoder in the network. This saves training time and improves scalability for future real-world deployments.
Tingfang Ji, vice president of engineering at Qualcomm Technologies, said: “AI is revolutionizing many aspects of our lives and society, and Qualcomm is leading the way in cutting-edge AI technology. As AI brings new capabilities, efficiency, and performance to multivendor wireless systems, The ecosystem must work together to enable wireless AI on a global scale. Working with Nokia on AI interoperability is an encouraging first step in defining a scalable global standard.”
Peter Vetter, Director of Core Research at Nokia's Bell Labs, said: “The potential for AI in communications is huge. Data-driven air interfaces can learn about the environment and improve performance and efficiency. Although AI is already being implemented in 5G and 5G-Advanced , its full potential will be realized in the 6G era.”
Resources and additional information
Blog: Scaling up AI/ML for cellular wireless access
Blog: Powering 5G Mobility with AI/ML – Advances paving the way to AI-native 6G
Video: 6G AI Native Air Interface Proof of Concept
This proof-of-concept demonstration will be on display at Mobile World Congress 2024 at Qualcomm's booth in Hall 3 Stand 3E10 and Nokia's booth in Hall 3 Stand 3A20.
About Nokia
At Nokia, we create technology that helps the world work together.
As a leader in B2B technology innovation, we are pioneering networks that sense, think, and act by leveraging our efforts across mobile, fixed, and cloud networks. Additionally, we create value through intellectual property and long-term research led by award-winning Nokia Bell Labs.
Service providers, enterprises and partners around the world trust Nokia today to deliver secure, reliable and sustainable networks, and to help them create the digital services and applications of the future.
Media inquiries
Nokia Communications, Corporate
Email: Press.Services@nokia.com
Follow us on social media
LinkedIn X Instagram Facebook YouTube
