Improved binary marine predator algorithm-based digital twin-assisted edge-computing offloading method

A vast number of mobile internet-of-things (IoT) devices are connected to the internet, and they constantly generate new computing tasks. Owing to an IoT device’s limited resources and restricted computing power, heavy computing tasks are generally offloaded to edge servers, where a digital twin (DT) of the IoT device is synchronized using the current information from the physical device, followed by completing the computing task and returning the results. Notably, this strategy must be completed within time constraints while ensuring that the device is not overtaxed by its computations. The problem is that offloading solution is not efficient enough to meet time and energy requirements for critical tasks. Simultaneously, the success rate of offloading solutions affects computing efficiency. To address these problems, this study formulates a mathematical model of energy consumption optimization under time constraints and proposes a binary problem-enabled marine predator algorithm. Simulation results demonstrate that our proposed method effectively meets deadlines while reducing energy consumption, even with an increasing number of users.