Nokia Bell Labs launches Natural-Language Networks technology

Natural-Language Networks
Sheila Zabeu -

November 08, 2023

Have you ever thought about configuring networks just by saying a few commands? Well, research carried out by Nokia Bell Labs has taken an important step towards making networks truly autonomous through what it calls Natural-Language Networks. The innovation will make it possible to operate networks through dialogues or even simple text commands. According to the researchers, the networks will understand the users’ intentions and will have the intelligence to act accordingly, autonomously.

Natural-Language Networks aim to eliminate the complexity of managing these environments and also enable faster responses to user demands. The technology uses Artificial Intelligence (AI) to facilitate network configuration and maintenance.

Natural language processing and AI and machine learning (ML) models interpret simple voice or text requests to automatically allocate the right combination of resources. In addition, natural language networks continuously learn from actions, optimising processes with each request and gaining more capacity to anticipate the demands of services and applications, adapting without any human intervention.

“It will no longer be necessary to read technical catalogs or complex API descriptions to configure networks. Instead, a simple line like ‘optimize the network at location X for service Y’ will work just fine. These requests can be used, for example, to configure a wireless network in a factory for robot automation or to optimize networks at an event that will receive a flood of social media posts,” explains Csaba Vulkan, research leader in network systems automation at Nokia Bell Labs.

Natural language-based networks are one component of a Nokia Bell Labs research initiative called UNEXT. The project, named after the legendary UNIX operating system invented by Bell Labs, is intended to redefine software and network systems in the same way that UNIX reshaped computing, making it easier to securely integrate anything into a network and turning the network itself into an operating system.

“Natural language networks are a preview of one of UNEXT‘s many features. The reduced complexity of network management fits perfectly with UNEXT’s goal of extending the reach of networked systems by eliminating barriers that prevent their interoperability,” explains Azimeh Sefidcon, head of network systems and security research at Nokia Bell Labs.

Nokia Bell Labs demonstrated the technology in a proof of concept for the first time at the Brooklyn 6G Summit, held from 31 October to 2 November in New York (USA).

More about UNEXT

An acronym for Unified Networking Experience, UNEXT is an interactive, self-managing operating system that treats each network element as an independent entity, breaking down interoperability barriers between these elements. In the opinion of Nokia Bell Labs, UNEXT will be a new stage in the evolution of network software, effectively turning the network itself into an operating system. The organization says that UNEXT and UNIX have more similarities than just their names and should have the same lasting impact on the computing world.

UNEXT is expected to be key in the 6G era, connecting many disparate systems needed for the applications of the future, from robots and autonomous vehicles to the metaverse and virtual reality. By bridging the gaps between suppliers, standards and operators, UNEXT will help merge digital and physical realities, predicts Nokia Bell Labs.

One of UNEXT’s main features is that it is an interactive, self-managing operating system for users and applications with resources distributed across heterogeneous and proprietary environments. UNEXT’s six focus areas are:

  1. Autonomous services: UNEXT will offer automation and autonomy in all phases of the service lifecycle.
  2. Data and knowledge services: The distributed intelligence in UNEXT will affect everything the network touches, automatically creating processes for agile data operations and machine learning, as well as knowledge and context generation.
  3. Network-application symbiosis: UNEXT will improve application performance and network efficiency using dynamic interactions.
  4. Extended computing services: UNEXT will orchestrate heterogeneous computing resources, from the cloud to the edge to devices.
  5. Decentralised environments: UNEXT will also securely and reliably orchestrate heterogeneous devices, components, entities and actors.
  6. E2E system dynamics: UNEXT will capture the desired properties, such as reliability, security and scalability, from end to end and automatically eliminate the undesired properties.