Enhancing NextG with ML-based Multi-User Resource Scheduling

Go back to programme

Enhancing NextG with ML-based Multi-User Resource Scheduling

  • Register

    * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

  • Massive MIMO base stations, with their large number of antennas, can use beamforming to communicate with multiple users simultaneously on the same time and frequency resource. However, when user channels are highly correlated, this advantage can be diminished, significantly impacting the spectral efficiency of multi-user beamforming. As users move within a network cell, their channels vary across both time and frequency, which in turn affects their correlation with other users, leading to different optimal groupings of users depending on the specific time-frequency resource. Thus, to maximize spectral efficiency while maintaining fairness among users, the base station must carefully schedule users during each transmission interval. 

    The goal of this challenge is to develop machine learning-based algorithms to address the multi-user beamforming scheduling problem in practical settings. Given the considerable overhead associated with constantly measuring user channels in mobile environments, an efficient solution may need to work with partial or even stale channel data to schedule users effectively in upcoming periods. Thus, machine learning algorithms can be ideal for managing incomplete or old channel data and deriving effective user scheduling strategies that aim to maximize network throughput and fairness. 

    In this webinar, we will discuss this challenge in greater detail. We will present a baseline solution that assumes full channel knowledge, outline the available datasets, and explain the evaluation criteria for the challenge. This discussion aims to foster innovation in developing robust solutions for multi-user scheduling in massive MIMO systems. 

    This live event includes a 20-minute networking event hosted on the AI for Good Neural Network. This is your opportunity to ask questions, interact with the panelists and participants and build connections with the AI for Good community.

    Share this session
    • Days
      Hours
      Min
      Sec
    • Watch

      Register

      * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

    • Start date
      13 May 2024 at 16:00 CEST Geneva | 10:00-11:00 EDT, New York | 22:00-23:00 CST, Beijing
    • End date
      13 May 2024 at 17:00 CEST Geneva | 10:00-11:00 EDT, New York | 22:00-23:00 CST, Beijing
    • Duration
      60 minutes (including 20 minutes networking)
    • Programme stream
    • Topics
    • UN SDGs

    Are you sure you want to remove this speaker?