ECC Symbol Ordering for Low-Buffer Serial Data Links
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data transmission methods face challenges in efficiently transporting error correction codes, leading to increased complexity and cost due to the need for complex FEC algorithms and large buffer sizes, especially in high-speed serial data links, where single bit errors can result in multiple bit errors.
Innovation Solution
A method where error correction codes (ECCs) are generated and transmitted ahead of the encoded data stream, allowing for symbol-based error correction before decoding, reducing the need for large buffers and complex FEC algorithms by enabling simple FEC algorithms like hamming codes to correct errors effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If forward error correction codes are transmitted after data encoding into symbols, then error correction capability is provided, but the receiver requires large buffer memory to store both data and parity symbols
Solution Approach 1:
The patent inverts the conventional transmission order by sending parity symbols before data symbols. This allows the receiver to have parity information available immediately when data arrives, eliminating the need to buffer both data and parity together. The receiver can process data symbols as they arrive using pre-received parity symbols, significantly reducing buffer memory requirements.
Solution Approach 2:
The patent implements preliminary action by generating and transmitting parity symbols before the corresponding data symbols. This preliminary provision of error correction information enables the receiver to correct errors in real-time as data arrives, without needing to accumulate a large buffer of both data and parity symbols first.
2Reliability
If complex FEC algorithms like reed-solomon or LDPC are used to correct byte errors, then error correction capability improves, but encoder and decoder complexity increases
Solution Approach 1:
The patent extracts the error correction function from the symbol decoding process and applies it at the byte level after decoding. Instead of using complex FEC algorithms during symbol-level processing, simple parity checking is performed on individual bytes after they are decoded from symbols. This separates the DC balancing function (handled by symbol encoding) from the error correction function (handled by byte-level parity checking), allowing simple algorithms to suffice.
3Reliability
If parity bits are generated for data bytes and encoded through DC balancing into symbols, then forward error correction is provided, but a single bit error in transmission can result in multiple bit errors in received data
Solution Approach 1:
The patent segments the error correction approach by applying parity checking at the byte level rather than treating entire blocks of symbols as single units. Each byte is independently checked against its corresponding parity byte after decoding. This segmentation limits error propagation to individual bytes rather than allowing errors to affect entire data blocks, reducing the impact of symbol-level transmission errors.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is provided a method of transporting error correction codes (ECCs) in a transmission stream, the method including encoding a data stream from a data source into data symbols, generating first ECCs from the data symbols, encoding the first ECCs into first error correction code (ECC) symbols, merging the data symbols and the first ECC symbols into the transmission stream, the first ECC symbols being merged before the data symbols into the transmission stream, and transmitting the merged transmission stream to a sink device via a communication link, the first ECC symbols being transmitted before the data symbols.