Working Group AI for Wireless Communications

A new WWRF Working Group

Introduction

The four dimensions describing intelligence are the abilities to:

  1. “perceive” (collect data information that can be rich, complex or subtle)
  2. “learn” (in a determined environment)
  3. “abstract” (enable the creation of new meanings) and
  4. “reason” (allow to plan and decide)

Within this context, Artificial Intelligence (AI) is commonly defined as the ability of a given cyber-physical system to perceive the environment, process and analyze the perceived information, and take a set of actions that maximize the likelihood of achieving a certain goal.

In the last few years, AI has been successfully applied to many domains including medical diagnosis, automatic pilots in avionics, and gaming, to name a few where AI performance has demonstrated capabilities beyond its human counterpart. More recently, AI is also being perceived as a potential technology solution to cope up with the increasing complexity of wireless communication system design and associated management. The former being induced by the extreme range of requirements for user experience, efficiency, and performance. Future wireless networks will require robust intelligent algorithms to adapt network protocols and perform network and resource management tasks for different services in different scenarios. For example, for network management and administration duties, AI will be mainly expected to make smart predictions, and to take improved decisions and actions that until now were mostly based on pre-defined rulesets obtained from experience accumulated by humans over the years.

When applied to the wireless communications domain, AI is referred to as Wireless Intelligence (WI). When fully operative, it is expected that WI will be transformative and pervasive. WI will be in all sub-systems within the wireless ecosystem; in devices, User Equipment, machines, cars or connected things.  The network elements themselves in Operation Support Systems (OSS), Business Support Systems (BSS), and application servers up on the cloud would also be impacted. WI is expected by the market to not only reduce operational expenditures (OPEX), but also to increase user quality of experience (QoE) as well as help the introduction of new value chains in an increasingly competitive and complex business environment.

WG Vision

The working group vision is to develop a white paper on Wireless Intelligence. This white paper, part of the Outlook series,  will outline the expectations of the industry and the market on WI and the industry commitments to shape, evolve and optimize its mobile broadband products portfolio at the right level to face the upcoming challenges on network design and management.

The working group will engage a critical mass of experts from industry and academia, in order to investigate/understand the trends, requirements, enablers and challenges in the adoption of AI techniques to manage and optimize wireless communication networks

Objectives

  • Study the technology requirements and research challenges related to Wireless Intelligence
  • Develop a white paper to announce the results of the study
  • Facilitate collaborations and consensus building among various experts and actors
  • Organize workshops to encourage dissemination and interactions
  • Follow/participate/contribute to research/technology/regulatory/standardization fora

Stakeholders

The following organizations have already expressed their interest to participate in this activity and/or have contributed to the special session organized WWRF39

  • CTTC
  • Huawei
  • Bell Mobility
  • CSIR
  • University of Piraeus
  • Kings College London

Other organizations will be invited to contribute to the new WG.

White Paper

WP1

  • Title to be determined
  • Timeline: September 2018 – March 2019
  • The white paper will be presented at the 42nd WWRF meeting in Spring 2019

Actions completed

  • Organization of a special session at WWRF39 entitled Wireless Artificial Intelligence.
  • Establishment of a special interest group (SIG-WAI) within WWRF with a dedicated mailing list
  • Hosted European launch of Wireless AI Alliance led by China Mobile, Huawei  and Alibaba
  • Extensive discussion and assessment of Huawei white paper as basis for our work

Actions planned

  • Development of new white paper
  • Presentation of the white paper results at the 42nd  WWRF meeting
  • Special session to be held at international conference (WCNC, EuCNC)

The WG chair is Tanbin (Huawei) and the proposed WP1 editor is Miquel Angel (CTTC).

The WG will run for 12 months (September 2018 – August 2019).