Split Rendering Connectivity for Media Processing Entities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MPEG Network Based Media Processing (NBMP) designs lack solutions for split rendering, device/platform resource description, connection characteristics between cloud nodes, and network element capability discovery.
Innovation Solution
Implementing a mechanism for split-rendering in NBMP standards, describing media source/sink device/platform capabilities, and extending APIs to discover cloud node and network capabilities, including parameters for connection characteristics like bandwidth and latency between media processing entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If media processing is performed entirely on cloud/edge resources, then processing capability is improved, but device complexity increases due to need for connectivity information exchange
Solution Approach 1:
The patent introduces connectivity information as an intermediary data structure that mediates between media processing entities. This connectivity information contains standardized connection characteristics (bandwidth, latency, reliability) that enable automatic workflow optimization without requiring complex direct negotiations between entities, thus resolving the contradiction between enhanced processing capability and device complexity
Solution Approach 2:
The patent transforms complex connection characteristics into quantifiable parameters (bandwidth, latency, reliability metrics). By changing the representation of connection data into standardized parameters, the system enables automated workflow optimization and resource allocation while reducing the complexity of direct entity communications
2Use of energy by moving object
If split rendering is implemented, then device power consumption is reduced, but system complexity increases due to workflow distribution
Solution Approach 1:
The patent applies segmentation by dividing the media processing workflow into multiple independent tasks that can be distributed across different entities (cloud, edge, device). Each task can be independently assigned and executed, enabling split rendering that reduces device power consumption while managing complexity through modular task decomposition
Solution Approach 2:
The patent implements dynamic workflow distribution where tasks can be dynamically assigned and reassigned based on connectivity information and system conditions. This dynamic approach optimizes energy consumption by adapting the workflow distribution to current network conditions while managing complexity through automated decision-making
3Productivity
If connectivity information is collected from all MPEs, then workflow optimization is improved, but information processing time increases
Solution Approach 1:
The patent applies preliminary action by collecting and storing connectivity information from all media processing entities in advance before workflow execution. This pre-collection of connectivity data (bandwidth, latency, reliability metrics) enables the workflow manager to quickly optimize task assignment without time-consuming real-time information gathering, thus improving productivity while reducing processing time
Data Source
AI summary
Systems and methods for split rendering of Network Based Media Processing (NBMP) are provided. A method includes deriving a media processing workflow; and obtaining connectivity information of at least one media processing entity (MPE) from among a plurality of MPEs, wherein the connectivity information describes a connection characteristic between at least two MPEs from among the plurality of MPEs; and assigning a plurality of workflow tasks of the media processing workflow among the plurality of MPEs based on the connectivity information.