Server Time Reservation for Periodic Real-Time Applications

To utilize edge and cloud in real-time industrial applications, communication with the edge and cloud servers should be predictable in timing. However, the predictability of offloading from device to servers cannot be guaranteed in an environment where multiple devices compete for the same edge and cloud resources due to potential server-side scheduling conflicts. To the best of our knowledge, the state-of-the-art lacks a technique for offloading real-time applications from multiple devices to a set of heterogeneous edge/cloud servers. To this end, this paper proposes a centralized resource reservation technique that enables the offloading of real-time applications to the edge and cloud in a predictable time-schedule. The proposed technique enables end-devices to request the server's time for offloadable real-time applications in advance, allowing a designated offloading server that guarantees the tasks' timely execution. Furthermore, the proposed technique is capable of optimizing the reservation scheduling strategy with the goal of minimizing the energy consumption of edge servers while meeting the stringent timing requirements of real-time applications. The results showed that the number of deadline satisfied jobs improved by 65%, and total energy consumption by 3%, compared to the second best algorithm among the ones that have been compared with the proposed algorithm when the number of jobs is changed.