Scalable Video Block Requests for Low-Latency Adaptive Streaming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current media streaming systems face challenges in providing uninterrupted playback and efficient bandwidth utilization, particularly in adapting to network and buffer conditions, leading to issues like playout stalling and excessive storage resource consumption.
Innovation Solution
A block-request streaming system that uses HTTP streaming with erasure codes and intelligent client procedures to manage media data requests, allowing for adaptive bit-rate delivery and efficient decoding, while minimizing latency and CPU usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional download or streaming models are used, then media delivery can be performed, but playout stalling occurs and user experience deteriorates
Solution Approach 1:
The patent segments media content into discrete blocks that can be independently requested and decoded. This segmentation allows the client to request specific blocks out of order based on network conditions and buffer status, preventing playout stalling while maintaining efficient bandwidth utilization. Each block is self-contained with appropriate headers and data, enabling independent processing.
Solution Approach 2:
The patent implements dynamic block request strategies where the client adapts its requesting behavior based on real-time network conditions and buffer status. The system dynamically adjusts which blocks to request, when to request them, and at what rate, optimizing both playout continuity and delivery efficiency under varying conditions.
2Manufacturing precision
If high quality media is delivered, then presentation quality improves, but bandwidth consumption increases
Solution Approach 1:
The patent enables dynamic parameter changes in media delivery by allowing the client to request blocks at different quality levels based on available bandwidth. The block structure supports variable quality parameters while maintaining consistent decoding efficiency, as each block is self-contained with appropriate headers that enable flexible quality adaptation without requiring complete re-encoding.
3Device complexity
If client requests media blocks in order, then decoding is simplified, but latency increases
Solution Approach 1:
The patent segments media into independent blocks with complete headers and metadata that enable self-contained decoding. This segmentation allows out-of-order requests and receipt without increasing decoding complexity, as each block contains all necessary information for independent processing. The client can request blocks in any order based on network conditions while maintaining efficient decoding.
Solution Approach 2:
The patent implements dynamic block request strategies where the client can adaptively request blocks out of order based on real-time network conditions and buffer status. This dynamic approach reduces playout latency by allowing the client to request anticipated blocks proactively while the player consumes currently available blocks, without requiring sequential receipt of all blocks.
4Reliability
If buffer size is increased to prevent stalling, then playout continuity improves, but storage resource consumption increases
Solution Approach 1:
The patent segments media into smaller, self-contained blocks that can be requested and delivered independently. This segmentation allows the client to maintain a smaller buffer by requesting specific blocks as needed rather than pre-loading large continuous segments. The block structure with complete headers enables efficient processing of smaller units, reducing the storage resources required in the buffer while maintaining playout continuity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In the current invention, a method of requesting media data at a client device using a scalable segment map is described. Thereby, the method comprises receiving the scalable segment map at the client device, the scalable segment map including metadata with one or more byte ranges corresponding to locations of layers of a media segment, wherein the media segment includes one or more blocks and is associated with a scalable video coded stream including multiple layers that are combinable to form one or more representations. Furthermore, the method comprises determining, by the client device, one or more layers of a block of the media segment to request, generating, by the client device, a request for the one or more layers of the block, the request including a byte range corresponding to a location of the one or more layers, wherein the byte range is determined from the metadata included in the segment map, transmitting, by the client device, the request for the one or more layers of the block, receiving, at the client device, the one or more layers of the block of media data and generating, by the client device, a media representation using the one or more layers of the block. A corresponding method at a media ingestion system is also described. Furthermore, the invention describes a related client device and a media ingestion system.