Processing Unit Selection for Low-Latency Mobile Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Services using remote processing units for rendering video frames are sensitive to delay, especially in mobile user equipment (UE), leading to degraded quality-of-experience (QoE) and issues like simulator sickness due to latency spikes during handovers.
Innovation Solution
A method for selecting the best processing unit (PU) from a set of available PUs based on performance and budget information, using a PU selection function (PSF) to minimize latency and prepare for handovers by pre-configuring PUs, ensuring optimal rendering even during mobility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If a remote processing unit is used to render video frames, then rendering power is improved, but latency increases
Solution Approach 1:
The system performs preliminary actions by pre-configuring and preparing the target processing unit before the actual handover occurs. This includes pre-establishing data flows, pre-loading necessary data, and pre-configuring rendering parameters so that when handover is triggered, the transition requires minimal additional time, thus reducing latency while maintaining access to powerful remote rendering resources
2Adaptability or versatility
If handover is performed during mobile UE movement, then adaptability is improved, but latency increases
Solution Approach 1:
The system proactively monitors UE mobility status and triggers handover preparation in advance based on predicted movement patterns or signal quality trends. By initiating the handover process before the UE actually moves out of the current cell coverage, the system reduces the effective handover latency while maintaining the ability to adapt to mobile conditions
Solution Approach 2:
The system dynamically adjusts handover parameters and timing based on real-time UE mobility status, signal conditions, and network load. This includes adjusting handover thresholds, preparation time windows, and resource allocation dynamically to optimize the balance between mobility adaptability and handover latency
3Adaptability or versatility
If remote GPU is located outside user premises, then resource availability is improved, but latency increases
Solution Approach 1:
The system implements local quality by caching frequently accessed rendering data, textures, and configuration information at edge locations or local buffers closer to the UE. This allows the system to maintain access to remote GPUs located outside user premises while reducing network latency for data retrieval and transmission by using locally cached resources when appropriate
Data Source
AI summary
A method (400) for selecting from a set of two or more available PUs (e.g. a set of GPUs or a set of TPUs) a PU to serve a UE (102) (i.e., select the “best” PU for the UE). The set of available PUs includes at least a first PU and a second PU. The method includes, for each PU included in the set of available PUs, obtaining (s402) performance information for the PU. The performance information for the first PU comprises a first time value indicating an estimated amount of time that it will take the first PU to produce a first result. The method also includes obtaining (s404) budget information for the UE, wherein the budget information comprises a threshold time value indicating a threshold amount of time. The method further includes selecting (s406) a PU from the set of available PUs based on the stored performance information and the budget information for the UE.