Nokia and Qualcomm Collaborate on Advancing AI-Interoperability Technology to Enhance Wireless Capacity and Performance at #MWC24
In a significant development, Qualcomm and Nokia Bell Labs have successfully addressed a key obstacle related to AI interoperability, which hinders the seamless collaboration among diverse artificial intelligence (AI) systems from various vendors across devices and networks.
Through the innovative application of sequential learning, distinct AI models can undergo independent training, enabling them to synchronize and cooperate without the necessity of sharing proprietary AI models, which serve as distinguishing factors for different vendors. This breakthrough has led to the creation of a prototype for the sequential training solution by Qualcomm and Nokia Bell Labs, set to be showcased at MWC24 in Barcelona.
Qualcomm Technologies and Nokia jointly announced the successful development of a cutting-edge technology prototype aimed at resolving interoperability challenges – a fundamental issue impeding the effective utilization of AI in wireless systems. By facilitating multi-vendor interoperability among different AI models utilized in networks and devices, future wireless systems are poised to leverage AI capabilities to enhance capacities, improve performance, and potentially reduce energy consumption.
The implementation of this groundbreaking technology into an AI proof-of-concept involved the utilization of a Nokia Bell Labs test network in conjunction with a Snapdragon 5G modem-RF system. The collaborative efforts of Qualcomm Technologies and Nokia Bell Labs will be demonstrated through this proof-of-concept at the upcoming Mobile World Congress 2024 in Barcelona.
The integration of AI and machine learning is set to revolutionize wireless communications by enabling AI models to optimize radio performance based on specific environmental conditions. However, for AI to realize its full potential within radio networks, it is imperative that different AI systems deployed across networks and devices effectively coordinate their operations to prevent conflicting actions.
Qualcomm Technologies and Nokia Bell Labs have introduced a novel approach to AI interoperability leveraging sequential learning. This methodology involves the sharing of connection-relevant data for AI training between the network and device without the necessity of sharing the actual AI model. Subsequently, independent encoders and decoders utilize this shared data to independently train the network AI and device AI. These independent encoders and decoders act as facilitators, enabling disparate AI systems to exchange knowledge seamlessly. Through sequential training, these diverse AI models can mutually reinforce each other, irrespective of the vendor. Furthermore, sequential training fosters compatibility between wireless handsets or devices from different vendors, each equipped with their unique encoder implementation, enabling them to seamlessly interact with a common decoder in the network’s base station. This approach not only streamlines training processes but also enhances scalability for future real-world deployments.
Tingfang Ji, VP of Engineering at Qualcomm Technologies, remarked: “AI has ushered in transformative changes across various domains, with Qualcomm at the forefront of edge AI technology. To harness the full potential of wireless AI capabilities on a global scale, it is imperative for the ecosystem to collaborate effectively. Our partnership with Nokia on AI interoperability marks a significant stride towards establishing a scalable global standard.”
Peter Vetter, President of Bell Labs Core Research at Nokia, added: “The potential for AI in communications is immense. A data-driven air interface can adapt and enhance its performance and efficiency based on environmental insights. While we are witnessing the integration of AI in 5G and 5G-Advanced networks, its true potential will be realized in the upcoming 6G era.”
Resources and Further Details
Blog: Scaling up AI/ML for cellular radio access
Blog: AI/ML enhanced mobility in 5G-Advanced paving the way for an AI-native 6G
Video: 6G AI native air interface proof of concept
The demonstration of this innovative proof-of-concept will be available for viewing at Mobile World Congress 2024, showcased at Qualcomm’s booth at Hall 3 Stand 3E10 and Nokia’s booth at Hall 3 Stand 3A20.
About Nokia
Nokia is committed to developing technology that fosters global collaboration and progress.
As a pioneering leader in B2B technology innovation, we are spearheading the creation of networks that possess the ability to perceive, analyze, and act upon information through our extensive work across mobile, fixed, and cloud networks. Additionally, we drive value through our intellectual property and long-term research endeavors, led by the esteemed Nokia Bell Labs.
Service providers, enterprises, and partners worldwide rely on Nokia to deliver secure, dependable, and sustainable networks today, collaborating with us to shape the digital services and applications of tomorrow.
Media Inquiries
Nokia Communications, Corporate
Email: Press.Services@ nokia.com
Follow us on social media:
LinkedIn X Instagram Facebook YouTube
