[Hardware] Ericsson teams with MIT on hardware design for cognitive networks

[Love Pulse]

 

Ericsson is exploring cognitive networks to reduce operational complexity and increase energy efficiency, through two major research projects in a new collaboration announced with the Massachusetts Institute of Technology.

The partners did not disclose the amount of funding or the length of the project, but an Ericsson spokesperson said work is already underway. The projects focus on long-term technical research to design advanced devices that could help power 5G and 6G networks in the future.

5G is paving the way for the full realization of the Internet of Things, Magnus Frodi, Head of Research at Ericsson, said in the announcement, citing “enormous amounts of small IoT devices and AI-powered cognitive networks” as “drivers for the next leap forward.”

 

“Working with the amazing teams at MIT, we hope to develop devices that make this possible,” Frodi stated.

The main focus is on research of lithium-based devices to enable neural computing. A December 2020 ZDNet article described Neural Computing, a brain-inspired model that uses less power to process artificial intelligence.

 

Cognitive networks that use AI for highly automated, secure, data-driven processes are gaining interest from Ericsson as the Swedish vendor looks to reduce complexity for operators who need to manage large mobile network environments with increasing levels of features.

 

Specifically, Ericsson Research and the Materials Research Laboratory at the Massachusetts Institute of Technology are collaborating to investigate new designs in Lithian chips that will enable neural computing "that delivers significantly more energy-efficient AI processing" than is currently available and could lead to Complete knowledge networks.

An Ericsson spokesperson explained that the scope of the research includes understanding both basic semiconductor technologies, as well as AI approaches, so the potential outcome should not be viewed as one specific circuit.

 

The project takes a holistic approach to mobile network research and the science of lithium is not limited to a specific subsystem and domain of the network.

The second focus is energy harvesting technologies, mainly applied to cost-effectively low-power IoT devices supported by the mobile network.

Anantha P. Chandrakasan, dean of the MIT School of Engineering, said in the announcement that the pair are tackling major technical challenges as energy-efficient devices continue to advance.

 

“By bringing together our knowledge and Ericsson’s experience in mobile technology, we aim to develop devices that will power exciting new AI applications at the edge, and make great strides in the next generation of mobile networks,” said Chandrakasan.

Ericsson will work with the MIT Electronics Research Laboratory to show how devices, such as massive amounts of sensors or other "power-free" devices, can directly harness energy from received radio signals to connect to the network. The duo will also look at systems design for how to take advantage of this lower power for simple tasks.

“Besides hardware device technology, our research will include how the mobile network (RAN and Core) can be designed to connect and control these devices,” an Ericsson spokesperson said by email.

 

It's a long-term effort, and he said the realistic realization of the results will likely be in the 6G horizon timeframe. 5G systems are still somewhat in the early stages of deployment, and while some early 6G efforts are underway, 2030 has been cited as an approximate timeline for the initial commercialization of 6G.

"However, the results can still be applied to improve existing, cutting-edge 5G systems and are not necessarily tied to a specific architectural generation," the spokesperson added.