Fairness-aware Latency Minimisation in Digital Twin-aided Edge Computing with Ultra-Reliable and Low-Latency Communications: A Distributed Optimisation Approach (Invited Paper)

The advanced development of communication technologies and computing platforms open opportunities to enable a wide range of time-sensitive services. However, designing an effective optimisation solution to deal with joint communication and computation resources is a challenging research direction. This paper addresses a fairness-aware latency minimisation problem in the digital twin (DT) aided edge computing with ultrareliable and low latency communications (URLLC). The optimal solution is obtained by jointly optimising various variables, namely, bandwidth allocation, transmit power, task offloading policies, and the processing rate of user equipment (UE) and edge server (ES). The formulated optimisation problem is highly complicated with many non-convex constraints and strong coupling variables. To deal with the problem, we propose a distributed optimisation solution based on the global consensus approach and the successive convex approximation framework (SCA). Selected numerical results are provided to validate the proposed solution in terms of minimise latency as well as improved fairness among all UEs in the DT network.

I. Introduction

Ultra-reliable low-latency communications (URLLC) and wireless edge computing are two significant technologies to enable a wide range of time-sensitive applications [1]. These technologies are the two key enablers for the development of the future wireless applications in terms of communication and computation aspects, respectively. According to the 3GPP Release 15, and the ongoing 3GPP Release 16, URLLC is being enhanced to achieve the stringent requirements of “five-nine” to “seven-nine” i.e. 99.99999% in reliability while ensuring the end-to-end latency on the order of 1 ms. In computing perspective, mobile edge computing (MEC) provides a powerful framework to leverage the computing capacities of nearby edge servers (ES) or fog-cloud servers to reduce to processing time of computational tasks. The studies of joint communication and computation resource management in these important technologies recently attracting much attention from the research community.