Hybrid Edge-Cloud Computational Offloading for XR Medical Applications

Abstract

Emerging Extended Reality (XR) applications, particularly in eHealth, offer new opportunities in digital healthcare, such as XR-assisted surgery. Nevertheless, the XR use case is set to extremely high standards to ensure safety, high quality of the medical service, and high user experience. Hence, workload-intence and latency-hungry XR applications force researchers to find ways to process data efficiently. Even though local processing advances data safety because no data is shared with the third-party device, it demands some computational capabilities, directly affecting the battery. Since Mobile Cloud Computing (MCC) or Mobile Edge Computing (MEC) provides a computationally rich server, the proposed hybrid model allows for controlling decision strategies and managing safety and response time according to the task. A hybrid offloading strategy decreases system latency by 77% compared to MCC, 60% improvements to local processing, and 11% enhancement to MEC offloading. The proposed hybrid system reduces the delay by providing computational resources closer to users but can be strained under high workloads.