The 5th generation (5G) and beyond network offers substantial promise as the ideal wireless technology to replace the existing inflexible wired connections in traditional factories of today. 5G network slicing allows for tailored allocation of resources to different network services, each with unique Quality of Service (QoS) requirements. This paper presents a novel solution for slice-aware radio resource allocation based on a convex optimisation control framework for applications in smart factory wireless networks. The proposed framework dynamically allocates minimum power and sub-channels to downlink mixed service type industrial users categorised into three slices: Capacity Limited (CL), Ultra Reliable Low Latency Communication (URLLC), and Time Sensitive (TS) slices. Given that the base station (BS) has limited transmission power, we enforce admission control by effectively relaxing the target rate constraints for current connections in the CL slice. This rate readjustment occurs whenever power consumption exceeds manageable levels. Simulation results show that our approach minimises power, allocates sub-channels to users, maintains slice isolation, and delivers QoS-specific communications to users in all the slices despite time-varying number of users and changing network conditions.

翻译：第五代（5G）及未来网络作为理想无线技术，在替代当前传统工厂中僵化的有线连接方面展现出巨大潜力。5G网络切片能够为具有独特服务质量（QoS）需求的不同网络服务实现资源的定制化分配。本文针对智能工厂无线网络应用，提出了一种基于凸优化控制框架的切片感知无线资源分配新方案。该框架将下行混合服务类型的工业用户划分为三类切片：容量受限（CL）切片、超可靠低时延通信（URLLC）切片和时敏（TS）切片，并为其动态分配最小功率和子信道。考虑到基站（BS）传输功率有限，我们通过有效放宽CL切片中当前连接的目标速率约束来实施接纳控制。当功耗超过可管理水平时，即触发此速率调整。仿真结果表明，尽管用户数量随时间变化且网络条件动态改变，我们的方法仍能最小化功率、为用户分配子信道、维持切片隔离性，并向所有切片中的用户提供满足特定QoS的通信服务。