Abstract
Next-generation wireless networks are expected to be ultra-dense in terms of users and be able to support delay-sensitive traffic. Multiple-user, multiple-input, multiple-output (MU-MIMO) offers a potential solution by multiplexing a large number of concurrent data streams in the spatial domain. The MU-MIMO user scheduling process involves allocating the users across the space, and time or frequency resources, such that a performance metric is maximized, and subject to specific (e.g., rate) constraints being met. However, user scheduling is a combinato-rial problem, making its optimal solution highly intricate. This paper introduces the orthonormal subspace alignment scheduling (OSAS) approach, designed to be scalable for use in highly-dense networks and optimized for low-latency communications. Its design prioritizes users that align to the standard orthonormal basis and features a novel pruning process that enhances the users' transmission rates. Comparative evaluations reveal that OSAS makes more efficient use of the available resources and offers higher performance than other state-of-the-art techniques, while exhibiting lower complexity.