Digital Twin Association for Securing Internet of Energy: Challenges and Solutions

Promoted by the pervasive intelligence synergy in 6G-empowered edge networks for the Internet of Energy (IoE), the Digital Twin (DT) equivalently mapping the physical system has been increasingly adopted for minimizing operational disruptions and facilitating cost-effective executions. However, it is crucial to efficiently associate constructed DT service models with edge servers in an optimal fashion. To this end, we present a three-layer architecture of DT association for securing IoE. Although this architecture provides promising solutions to DT association, it also poses three challenges. To tackle these challenges, we present an utility-perception-driven enabler as a policy to find the association decisions while balancing the immediate response needs and the sustainable development needs in the whole life cycles of DT models. Particularly, we present an approach of short-long term utility perception with consideration of both short-term utility and long-term utility for DT associations. Through transforming the optimization of maximizing short- and long-term utilities as a Markov Decision Process, we design multi-agent deep reinforcement learning to achieve an efficient solution. Moreover, we verify the effectiveness and efficiency of the proposed solution by the experiments. Finally, we present an outlook on open issues of using DT in IoE.