| dc.description.abstract | For future networks, it is highly demanding to satisfy a wide range of time-sensitive and computation-intensive services. This is a very challenging task, since it requires a combination of aspects from information, communication, and computation in order to establish a digital representation of the real network environment. This paper introduces a fairness-aware latency minimisation (FALM) framework in the digital twin (DT) aided edge computing with ultra-reliable and low latency communications (URLLC), which jointly optimises various communication and computation parameters, namely, bandwidth allocation, transmission power, task offloading portions, and processing rate of user equipment (UEs) and edge servers (ESs). The formulated problem is highly complicated, due to non-convex constraints and strong coupling among optimisation variables. To deal with this problem, we develop both centralised and distributed optimisation approaches. In particular, we first resort to the successive convex approximation (SCA) method to develop a low-complexity iterative algorithm and solve the problem in a centralised manner. Combining tools from SCA and the alternating direction method of multipliers (ADMM), we develop an efficient distributed solution with parallel computation processing at ESs under global consensus in each iteration and strong theoretical performance guaranteed. Numerical results are provided to validate the proposed solutions in terms of convergence speed and overall latency as well as improving fairness among all UEs. | en_US |
