Joint Throughput and Users Maximization in Digital Twin (DT) Assisted Mobile Edge Computing Networks

Abstract

Considering the rapid growth in user count and increasing demand for higher data rates, we need to optimize network strategies to accommodate more users and increase the network throughput. Recently, a novel concept i-e., Digital Twin (DT) has re ceived considerable attention in academia and industry due to its ability to monitor, analyze, and optimize physical systems. It connects the physical systems with the virtual spaces. Moreover, it helps to optimize decision-making. Therefore, this thesis aims to accommodate the maximum IoT nodes and increase the network through put simultaneously by considering the IoT node association and power allocation in the Digital Twin-assisted Mobile Edge Computing (MEC) network. The DT of each cloudlet is considered here which helps to optimize the power allocation and IoT node association. Generally, the DT technology helps in making more accurate and opti mized decisions in the MEC network by creating real-time digital representations of physical objects. This work formulates the optimization problem as a Mixed Integer Nonlinear Programming Problem (MINLP). To solve the proposed problem, the Outer Approximation Algorithm (OAA) is used due to its lesser complexity. The proposed algorithm’s convergence, effectiveness, and lesser complexity leads to ξ-optimal solu tion = 10−3 , achieved using standard problem solvers. The simulation results in terms of associated IoT nodes and the network’s throughput demonstrate the proposed ap proach’s effectiveness. The results show that the network throughput and associated IoT nodes are increasing with an increase in considered IoT nodes. Also, when the re viii sults of the DT-assisted MEC network are compared with the MEC network, it shows that by using the proposed approach, the DT-assisted MEC network has greater net work throughput than the MEC network. Moreover, the proposed DT-assisted MEC network can facilitate more IoT nodes as compared to the MEC network.