Summary The mixed reality is a new environment combining virtual reality (VR) with augmented reality (AR). The technologies related to mixed reality is growing with the research of visual computer science, however it is still uncomfortable user-experience because of bulky hardware. To improve user experience in mixed reality, we are now studying the low-latency integrated information processing with 5G edge-cloud computing and multi-tenant interaction scheduling technology.
1. Accelerate the multiple task processing VR/AR interaction with integrated informationTo accelerate the multiple task processing VR/AR interaction with integrated information, in 5G network, we consider offload processing with heterogeneous hardware acceleration in edge-cloud computing environment. This technology contains below features:
- Distributed parallel processing for low-latency and multi-tenant VR/AR interaction on heterogeneous hardware.
- Resource orchestration of hardware-accelerator based on VR/AR-operations specific computation.
- Development of service profiling information-based analysis technology for low-latency.
- Monitoring function to recognize GPU/FPGA occupancy status information-based resource status.
Figure 1. Accelerate the multiple task processing VR/AR interaction with integrated information2. Offload technology based on edge-cloud computingTo support offload technology based on edge-cloud computing, we are now researching cooperative scheduling synchronization. When supposing multi-tenant environment, a scientific knowledge of the multi-access offload and synchronized load balance is required and its characteristics have below followings:
- Offloading work to edge and performing synchronization in edge environment. parallel processing for low-latency and multi-tenant VR/AR interaction on heterogeneous hardware.
- Development of multiple offloading support technology that can efficiently handle collaboration synchronization among multiple participants, taking into account edge environment characteristics.
- Provide reliable service by controlling each offloading workload of multiple participants fluidly for dynamic resource environments
Figure 2. Offload technology based on edge-cloud computing