Skip to main content
Aug 08 2018

AI at the edge of innovation

As massive and rapidly changing connectivity and coverage demands are placed on 5G networks by billions of IoT devices and a growing array of consumer services, networks will need to predict and respond to those changes in real-time. 5G networks need to be so intelligent that it can nurture itself. AI needs to be embedded in the architecture to provide continuous, real-time, complex cognitive processing to better shape and size network resources when and where needed.

Ever since we announced ReefShark at Mobile World Congress 2018, a revolutionary chipset that features cutting edge silicon design and process technologies and enables integration of Artificial Intelligence (AI), we received significant interest from Mobile Operators around the world who want to understand the benefits of AI for their 5G networks. Nokia recently signed an MoU with China Mobile to research radio resource management and traffic prediction using AI and machine-learning in 5G RAN network. We also launched Nokia AirFrame Open Edge Server, the industry’s first Edge Cloud data center solution to meet the stringent and diverse low-latency data processing demands and advanced applications for consumers and industries.

Recently at the Google Cloud Next '18 event, Google announced Nokia as a partner to evaluate Google’s new Edge TPU (Tensor Processing Unit) and Cloud IoT Edge software for possible use in Nokia product platforms. Edge TPU is Google’s purpose-built ASIC chip designed to run TensorFlow Lite ML models at the edge. The Edge TPU has specialized hardware to perform dense vector and matrix computations and is optimized for Machine Learning applications in terms of footprint, performance and power. As a part of the evaluation, the Edge TPU would be seamlessly integrated as a companion chip to extend ReefShark capabilities for Machine Learning. The evaluation leverages Nokia 5G Future X architecture including Nokia ReefShark-powered AirScale Cloud RAN and AirFrame open edge server.

The way in which applications, networks and services will be built in the future is changing rapidly. As mentioned by Nokia CEO Rajeev Suri in his blog on Partnering to win in the 4th Industrial Revolution, “By leveraging one another’s unique strengths, we can go farther and faster to develop breakthrough, truly interconnected solutions and seize new market opportunities.”, collaboration is the most effective way to accelerate innovation, and openness is a guiding principle in a complex environment where customers, collaborators and communities interact.  We are looking forward to getting under the hood of this powerful Edge TPU and understanding if and how we can use it in our own developments. The evaluation would focus on the overall hardware capability and platform performance post integration of Edge TPU on Nokia product platforms.

Intelligence at the edge will help our customers transform their networks, helping to manage the complexity of operations and create many new business opportunities for them.

Stay tuned for more updates on Nokia’s AI innovations through our website.

Share your thoughts on this topic by joining the Twitter discussion with @nokia and @nokianetworks using #5G #AI #cloud #machinelearning #innovation

About Tero Rissa

Tero is a member of Nokia’s mobile networks team and has 20 years of experience of scaling new technologies from research and concept engineering into commercial products. Outside the office, you’ll most likely find Tero outdoors, or cooking something overly elaborate.

Tweet us @nokianetworks