Multi-Channel Transport Scheduler for Extended Reality in Distributed Computing

With the advancements in 5G technology, ultra-low latency communications is possible which enables a new realm of experiences for users in the form of augmented reality (AR), virtual reality (VR), and extended reality (XR). To enable these new experiences, XR devices are being developed. However in present scenario, these devices are not able to perform high complex computation (on-device). Hence, these devices rely on companion computing devices (such as edge) to support complex computations. This results in a large amount of data being transferred from the XR device to the companion device at regular intervals. If multiple applications require the services of the companion device, then data from multiple sensors has to be transmitted to companion devices. To transmit every data together can flood the network, leading to network congestion, delay for latency-sensitive data, and deterioration of user experience at XR device. To address these issues, we propose a Multi-channel Transport (MCT) Scheduler mechanism that prioritizes data packets based on the priorities set by the applications at XR device. The MCT Scheduler mechanism uses a dual tag mechanism comprising an on-demand tag and a content tag to categorize data into multiple channels and prioritize the requirements of the applications. We conduct extensive experiments to validate the proposed MCT Scheduler mechanism performance with respect to the state of art in real-time environment.