Cloud Streaming Frame Segmentation for Resource Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud streaming services face challenges in efficiently providing multimedia content to a large number of concurrent users due to high resource consumption, particularly in legacy and thin client Set-Top Boxes, as existing methods struggle to manage resource overload and differentiate between dynamic and static content areas.
Innovation Solution
The method involves segmenting frames into dynamic and static areas, encoding the dynamic areas using intra mode and static areas using skip mode, and merging these streams to reduce network resource consumption and accommodate more concurrent users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cloud streaming services encode and transmit entire frames to all users, then all users can receive complete video content, but network resource consumption increases significantly
Solution Approach 1:
The patent segments video frames into multiple tiles (e.g., 4x4 grid) and allows selective transmission of only those tiles that contain dynamic content or are currently visible to the user. This segmentation enables the system to transmit a fraction of the total frame data while maintaining video content completeness for the user's view area.
Solution Approach 2:
The patent applies different encoding and transmission qualities to different regions of the frame based on their importance. High-priority regions (e.g., current display area, motion areas) are transmitted with full quality, while low-priority regions are either compressed more aggressively or omitted entirely, optimizing network resource usage while maintaining perceived video quality.
2Productivity
If the Cloud Streaming Server processes multiple video playback tasks simultaneously, then more concurrent users can be served, but resource overload occurs making it difficult to smoothly provide service
Solution Approach 1:
The patent segments the video processing task into independent tile-level operations that can be processed in parallel. By dividing frame encoding and transmission into smaller independent units (tiles), the server can distribute processing across multiple cores or nodes, increasing throughput and accommodating more concurrent users while maintaining service smoothness through efficient resource utilization.
Solution Approach 2:
The patent performs preliminary encoding of video frames into tile-based formats and pre-processes content during off-peak periods. This preliminary action prepares data in advance for rapid delivery during peak usage, enabling the server to handle more concurrent users without compromising service smoothness when demand is high.
3Adaptability or versatility
If legacy or thin client Set-Top Boxes run resource-intensive user interface scenarios, then user interface functionality is maintained, but device performance deteriorates due to limited resources
Solution Approach 1:
The patent extracts complex video processing and rendering operations from the Set-Top Box to the Cloud Streaming Server. By moving the computational burden of video decoding, tile generation, and rendering to the cloud, legacy and thin client STBs can maintain full user interface functionality while experiencing minimal performance degradation, as they only need to handle simple tile composition and display.
4Manufacturing precision
If the entire frame is encoded using intra mode for dynamic areas, then motion content quality is improved, but encoding complexity and time increase
Solution Approach 1:
The patent applies intra-mode encoding selectively only to tiles or regions containing motion or dynamic content, while static regions use more efficient prediction modes. This localized application of high-quality encoding to only where needed maintains motion content quality while significantly reducing overall encoding complexity and processing time compared to encoding the entire frame with intra mode.
Data Source
AI summary
Disclosed herein are a method and apparatus for providing a cloud streaming service. The method for providing a cloud streaming service includes segmenting the frame to be provided through cloud streaming into a first area and a second area, generating a first elementary stream by encoding the first area, generating a second elementary stream by encoding the second area, and merging the first elementary stream and the second elementary stream and transmitting a merged stream based on a request from a user.