A Framework to Maximize the Capacity of 5G Systems for Ultra-Reliable Low-Latency Communications

Abstract

Providing Ultra-Reliable Low-Latency Communications (URLLC) is one of the key challenges for 5G systems, which requires the redesign of both radio access technologies and the architectures. While 3GPP and the research community focus mainly on radio low layer functionality and architectural changes that enable URLLC, this article proposes a novel cross-layer framework that extends the already standardized 5G solutions to support a very massive number of users/flows in the system. The framework consists of a scheduler, a congestion avoidance algorithm, and a new traffic management algorithm, which together achieve the maximum capacity of 5G systems for URLLC, i.e., they maximize the load at which the reliability and latency requirements are satisfied for at least 99 percent of the users. Extensive performance evaluation results demonstrate that the proposed framework achieves the network capacity very close to the upper bound.