Distributed XR Compute Allocation for Latency and Power Tradeoffs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face challenges in efficiently managing extended reality (XR) compute locations for XR devices, UEs, and application servers, leading to suboptimal resource allocation and performance in XR applications.
Innovation Solution
The implementation of a dynamic distributed XR compute system that determines and selectively transmits XR compute locations based on various parameters, allowing for flexible and optimized resource allocation among XR devices, UEs, and application servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If XR compute is centralized at application servers, then processing power and resource availability are improved, but network latency and bandwidth consumption increase
Solution Approach 1:
The patent segments XR compute tasks into different components and distributes them across multiple locations (XR device, UE, network node, application server) based on task type and current system state. This allows critical rendering tasks to be performed locally while less time-sensitive tasks are handled remotely, resolving the contradiction between centralized processing power and distributed latency.
Solution Approach 2:
The system dynamically determines compute location based on real-time parameters including device capabilities, network conditions, and task characteristics. This dynamic allocation allows the system to shift compute tasks between local and remote locations optimally, balancing processing power availability against network latency for different XR applications and usage scenarios.
2Loss of time
If XR compute is distributed to local devices, then network latency is reduced, but device power consumption and processing limitations increase
Solution Approach 1:
The patent implements partial local processing where only the necessary portion of XR compute tasks is performed at the XR device or UE, while other portions are offloaded to application servers. This partial action approach reduces device power consumption by avoiding full local processing while still achieving low latency for time-critical rendering operations.
Solution Approach 2:
The system changes operational parameters (compute location, task allocation) based on device state, network conditions, and task requirements. When device power is sufficient and latency is critical, more compute is performed locally; when power is constrained or tasks are less time-sensitive, compute is shifted to servers, optimizing the balance between power consumption and latency.
3Productivity
If compute resources are dynamically allocated, then resource utilization efficiency is improved, but system complexity increases
Solution Approach 1:
The patent creates a universal compute determination mechanism that can handle multiple XR device types, task categories, and network scenarios through a single framework. This multi-functional approach improves resource utilization efficiency across diverse situations while avoiding the need for separate complex systems for each specific case.
Solution Approach 2:
The system employs feedback loops where compute location determinations are made based on current system state, executed, and then adjusted based on performance outcomes. This feedback mechanism enables dynamic resource allocation that adapts to changing conditions, improving overall resource utilization efficiency while the automated feedback control manages system complexity.
Data Source
AI summary
Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may determine, based at least in part on one or more parameters, an extended reality (XR) compute location for XR data associated with an XR device that is associated with the UE, wherein the XR compute location corresponds to the UE, the XR device, or an application server associated with the XR data. The UE may selectively transmit an indication of the XR compute location to a network node. Numerous other aspects are described.