A RAN Resource Slicing Mechanism for Multiplexing of eMBB and URLLC Services in OFDMA Based 5G Wireless Networks

Praveenkumar Korrai,Eva Lagunas,Bjorn Ottersten,Ashok Bandi,Symeon Chatzinotas,Shree Krishna Sharma

doi:10.1109/access.2020.2977773

Abstract

Enhanced mobile broadband (eMBB) and ultra-reliable and low-latency communications (URLLC) are the two main expected services in the next generation of wireless networks. Accommodation of these two services on the same wireless infrastructure leads to a challenging resource allocation problem due to their heterogeneous specifications. To address this problem, slicing has emerged as an architecture that enables a logical network with specific radio access functionality to each of the supported services on the same network infrastructure. The allocation of radio resources to each slice according to their requirements is a fundamental part of the network slicing that is usually executed at the radio access network (RAN). In this work, we formulate the RAN resource allocation problem as a sum-rate maximization problem subject to the orthogonality constraint (i.e., service isolation), latency-related constraint and minimum rate constraint while maintaining the reliability constraint with the incorporation of adaptive modulation and coding (AMC). However, the formulated problem is not mathematically tractable due to the presence of a step-wise function associated with the AMC and a binary assignment variable. Therefore, to solve the proposed optimization problem, first, we relax the mathematical intractability of AMC by using an approximation of the non-linear AMC achievable throughput, and next, the binary constraint is relaxed to a box constraint by using the penalized reformulation of the problem. The result of the above two-step procedure provides a close-to-optimal solution to the original optimization problem. Furthermore, to ease the complexity of the optimization-based scheduling algorithm, a low-complexity heuristic scheduling scheme is proposed for the efficient multiplexing of URLLC and eMBB services. Finally, the effectiveness of the proposed optimization and heuristic schemes is illustrated through extensive numerical simulations.

Highlights

The third and fourth generations (3G and 4G) of wireless networks have already revolutionized social behaviors through empowering the generalization of social networking on wireless mobile devices [2]
Rates to further improve the current mobile services such as high definition (HD) video and virtual reality (VR); ultra-reliable and low latency communications (URLLC) service concentrates on supporting low-latency transmissions of small packets with high reliability and it covers applications such as autonomous vehicles, industrial automation, and vehicular communications; mMTC supports the services that connect a massive number of devices where each device transmits small data packets intermittently and it covers the applications like smart cities
The resource allocation problem was formulated as an AMC based resource optimization problem to maximize the sum-rate of the total network while satisfying the heterogeneous requirements of the users from two services

Summary

INTRODUCTION

The third and fourth generations (3G and 4G) of wireless networks have already revolutionized social behaviors through empowering the generalization of social networking on wireless mobile devices [2]. None of works have targeted a design of the RAN resource slicing mechanism that efficiently optimize the currently available LTE standard radio resources (i.e., 0.5ms each transmission time interval (TTI), 1ms sub-frame and 10ms frame) between eMBB and URLLC services according to their isolation constraints and QoS requirements such as latency, reliability and minimum data rate. CONTRIBUTIONS In the above context, different from the existing works, in this work, we propose a RAN resource slicing technique for the efficient multiplexing of eMBB and URLLC services in wireless networks by considering an AMC scheme This resource slicing problem is formulated as an optimization problem to maximize the sum rate of all users, while satisfying the isolation constraint and stringent QoS constraints of the users such as latency and reliability.

SYSTEM MODEL

NUMERICAL EVALUATIONS

9: Find the highest SNR received RB:

24: Compute the delivered data of eMBB users

Findings

CONCLUSION