Broadcast Packet FEC Coding With Segmented Source Symbol Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In broadcasting and communication systems, data loss due to network congestion, especially with high-capacity content like HD and UHD, leads to user inconveniences such as audio degradation, video distortion, and file loss, as existing Forward Error Correction (FEC) methods require lengthy codes that increase complexity and decoding latency, and are inefficient in reducing zero-padded data.
Innovation Solution
A method for configuring a source block with variable-length data packets and adding repair information, dividing the block into regions to reduce zero padding and improve error correction, using signaling information to determine the positions of erased symbols for efficient FEC decoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of substance
If the desired length of source packets is reduced to minimize zero padding, then the amount of zero padding is reduced, but the number of symbols constituting the source block significantly increases, requiring FEC codes of very long length which increases decoding latency and complexity
Solution Approach 1:
The source block is divided into multiple sub-blocks, each processed independently with separate FEC encoding. This segmentation allows the system to maintain small packet lengths (reducing zero padding) while keeping the FEC code length manageable (reducing decoding latency and complexity). The receiver can decode multiple smaller sub-blocks in parallel rather than processing one large block sequentially.
Solution Approach 2:
The patent introduces a two-dimensional structure to the source block organization, arranging source packets in a grid format with rows and columns. This dimensional reorganization allows flexible grouping of packets into sub-blocks, enabling the system to optimize both packet length and code length independently by adjusting the row/column configuration, thus resolving the contradiction between minimizing zero padding and maintaining acceptable decoding latency.
2Loss of substance
If FEC codes of very long length are used to accommodate reduced packet lengths, then zero padding is minimized, but the complexity during Maximum log-Likelihood (ML) decoding significantly increases
Solution Approach 1:
By dividing the source block into multiple smaller sub-blocks, each with its own FEC code, the patent avoids the need for very long FEC codes. This segmentation directly reduces decoding complexity since ML decoding complexity grows exponentially with code length. Each sub-block can be decoded independently with simpler, shorter codes while still providing comprehensive error protection for the entire data set.
Solution Approach 2:
Instead of applying FEC encoding to the entire large source block at once (excessive action), the patent applies FEC encoding partially to smaller sub-blocks. This partial encoding approach reduces the computational burden on the decoder while maintaining sufficient error correction capability for each individual sub-block, thus reducing overall decoding complexity.
3Stability of the object's composition
If zero padding is increased to maintain source block structure, then packet alignment is improved, but the number of encoding symbols increases, generating and transmitting parity for unnecessary information
Solution Approach 1:
The patent segments the source block into multiple sub-blocks, allowing each sub-block to have its own optimized structure with minimal or no zero padding. This segmentation enables the system to maintain structural integrity at the sub-block level without requiring excessive zero padding at the overall block level, thereby reducing the generation and transmission of parity for unnecessary padded information.
Solution Approach 2:
Different sub-blocks can have different structures and padding requirements based on their specific content needs. This local optimization allows each sub-block to be configured efficiently for its particular data characteristics, rather than forcing a uniform structure with excessive padding across the entire source block. The receiver can process each sub-block with its optimal structure, improving overall transmission efficiency.
Data Source
Figure 1a~1b
Figure 2~3
Figure 4~5
AI summary
A method and apparatus for forward error correction, FEC, coding is provided. The method includes generating, based on a plurality of first source packets and a number of subdivided source symbols included in one source symbol, a source symbol block including a plurality of source symbols, wherein each of the plurality of source symbols includes one or more subdivided source symbols; generating, based on the source symbol block, the one or more first repair symbols by using FEC coding; generating, based on a source packet from among the plurality of first source packets, a second source packet; generating, based on one or more second repair symbols from among one or more first repair symbols, a repair packet; and transmitting the second source packet and the repair packet; wherein a size of each of the plurality of source symbols included in the source symbol block is same each other.