Performance Evaluation of Dynamic Computation Offloading Capability for Industrial Wearables

Abstract

Extended Reality (XR) is a disruptive technology that will play an essential role in future society by creating an immersive human-machine interface. For their mass adoption, XR head-mounted devices have to be made light and sleek in design, which may require distributed computing capabilities, where high-end devices wirelessly offload computational tasks to the accompanying processing units. To satisfy demanding wireless connectivity requirements of the emerging XR applications, the devices are expected to rely on radio technologies that operate in millimeter-wave (mmWave) frequency range and require directional transmission. In this paper, we evaluate a dynamic system of distributed 3D wearable networks operating in the mmWave band. We provide closed-form expressions for the session drop probability, the mean number of sessions that can run simultaneously, and its lower bound, which aid in understanding the impact of different parameters on the coexistence of dense directional wearable networks in the 3D space.