Layered Media Packetization for Congestion-Adaptive Quality Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media compression methods, such as JPEG, are not designed to adapt dynamically to network congestion, leading to dropped packets and inefficient resource usage when network conditions change, resulting in incomplete or delayed media delivery.
Innovation Solution
Partition media data into blocks, apply a transform to generate quantization coefficients, and rearrange them into quality layers based on frequency, allowing for dynamic reduction in quality during network congestion by prioritizing the most important information for transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If media data is compressed to a predetermined size at encoding time, then the data size is fixed and can be sent to the user, but the static size cannot support responsive actions to changes in network status leading to packet drops
Solution Approach 1:
The media data packet is segmented into quality layers with decreasing priority based on frequency information. Each layer contains quantization coefficients that can be independently transmitted or dropped. This segmentation allows the network to selectively transmit only the most important layers during congestion, ensuring basic media delivery while maintaining adaptability to network conditions.
Solution Approach 2:
The patent introduces dynamic adaptability by organizing media data into hierarchical quality layers that can be flexibly transmitted based on real-time network conditions. Instead of a fixed-size packet, the system can dynamically adjust which layers are transmitted, allowing responsive adaptation to network status changes while maintaining reliable delivery of critical information.
2Productivity
If the entire packet is dropped due to network congestion, then higher priority traffic can be managed, but the media data cannot be delivered and retransmission is required
Solution Approach 1:
The patent extracts and prioritizes the most important quantization coefficients into separate quality layers that can be transmitted independently. During network congestion, only the essential layers are transmitted, effectively taking out the critical information from the full packet. This allows the network to manage congestion by selecting which layers to transmit while ensuring that core media data is always delivered.
Solution Approach 2:
The system changes the transmission parameter from fixed-size packets to variable-size layered packets. By organizing data into quality layers with decreasing priority, the system can dynamically adjust the amount of data transmitted based on network conditions, changing the effective packet size and composition to balance throughput efficiency with reliable media delivery.
3Quantity of substance
If JPEG compression is applied to reduce data size, then the image can be transmitted efficiently, but the compressed data is arranged by blocks not by priority making packet wash operations incompatible
Solution Approach 1:
The patent segments the JPEG-compressed quantization coefficients into quality layers based on their priority and impact on media quality. Instead of maintaining the traditional block-based arrangement, the coefficients are reorganized into layers where each layer contains coefficients of similar importance. This segmentation enables packet wash operations to selectively remove lower-priority layers while preserving critical data.
Solution Approach 2:
The patent introduces a new organizational dimension for compressed media data by arranging coefficients in quality layers instead of traditional block structures. This dimensional reorganization adds a priority-based hierarchy that enables packet wash operations to function with compressed media, allowing network nodes to make intelligent decisions about which data to retain or discard during congestion.
Data Source
AI summary
A media data coding mechanism is disclosed. The mechanism includes partitioning media data into a plurality of blocks. A transform is applied to the blocks to obtain a plurality of quantization coefficients. The quantization coefficients are sorted into quality layers of decreasing priority based on frequency, wherein each subsequent layer includes data to incrementally increase quality of a reconstructed media data. Quantization coefficients for all blocks are positioned in a media data packet according to quality layer in order of decreasing priority. The media data packet is stored.