The US Department of Energy has revealed that they are financing $1.2m to significantly advance 5G technology with artificial intelligence (AI).
The considerable funding, which will be invested over a three-year period, has been awarded to the Argonne National Laboratory and Northwestern University, two pioneering institutions of 5G technology that are well-equipped and now well-funded to innovate the next generation of 5G communications.
The objective of the endeavour is to develop novel AI for 5G-enabled edge computing, a major evolution of 5G technology that will aid in optimising the quantity of scientific data that can be transferred over 5G networks. These refinements in 5G communications will have an array of benefits to multiple fields, such as atmospheric and environmental science.
Pete Beckman, the Co-Director of Northwestern Argonne Institute of Science and Engineering, said: “By combining the newest breakthroughs in AI learning and low-power AI accelerators, our 5G research project will give scientists new capabilities for adaptive, responsive, and smart distributed sensor networks.”
Improvements of 5G technology on communications
The 4G networks that have been used globally for many years are designed to transfer data between a data centre and a device. However, the new 5G technology that is being installed around the world supports computing within the network from edge devices, such as weather sensors and cameras, to the radio towers between the edge and the cloud.
4G networks transfer data, whereas 5G networks support computation at each phase along the digital continuum. For example, a sensor that a farmer can use to measure soil moisture or a micro radar unit that tracks and estimates the direction of a moving storm, of which data is then transmitted to a radio tower and aggregated before it is sent to the data centre.
How will the funding be utilised?
The Argonne National Laboratory has an illustrious history in the field of 5G technology, having already produced Waggle – an edge sensing platform that has led the industry of AI-enabled edge sensing for years. Waggle was employed by the Array of Things project at the University of Chicago, an investigation into urban dynamics that implemented over 100 intelligent sensors across the city.
Northwestern University is employing the latest iteration of Waggle in their Sage project, a venture that will examine ecosystems and aim to predict wildfires in California, Colorado, and Oregon. The novel Argonne edge platform will be crucial to enabling the technology for the laboratory’s 5G research.
Beckman said: “Wildebeest, the name of our 5G research project, will give scientists new capabilities for adaptive, responsive and smart distributed sensor networks. Wildebeest will support advanced AI that can ‘migrate’ across the digital continuum to allow us to start automatically optimising sensors and instruments in the field to report the most scientifically valuable data in real-time.”
The team is incredibly excited about their new network’s potential capabilities, as 5G is far superior in terms of versatility to the conventional alternative. In 4G technology, the provider controls the bandwidth, whereas, in 5G technology, individual users can make bespoke adjustment requests to the network, such as bandwidth and latency.
It is also possible to construct 5G wireless infrastructure in remote areas where 4G is otherwise inaccessible, with users able to set up their own small devices that can provide wireless service for several miles. This could be utilised for remote ecosystem monitoring in tribal communities.
Argonne will leverage AI to make real-time decisions, allowing them to discover new methods for automated radio management dependent on localised atmospheric conditions and the specific scientific goals for the distributed sensing. The project will see Argonne create innovative AI algorithms across the digital continuum to enhance distributed scientific instruments that utilise the network. The team will employ AI processing units in Waggle to create a prototype 5G testbed where AI workloads are able to migrate from edge to cloud, based on their scientific needs.
Nicola Ferrier, an Argonne scientist on the Wildebeest project and lead for the AI algorithm research, said: “With Wildebeest, distributed sensor networks will take a step toward autonomous sensing, automatically finding the most important data and using edge computing to respond hundreds of times faster.”