Flexible matrix-based FEC uses parity tags and dynamic parity group lengths to recover missing blocks under variable satellite data rates.
Intra-codeword interleaving and parity-check matrix rearrangement improve SD-LDPC BER under burst-mode upstream errors in 50G-PON.
Global and local syndrome decoding lets a constacyclic ECC decoder correct burst and multi-bit memory errors with less reliability loss.
Grouped and shifting H-matrix ECC helps memory systems correct multi-bit errors without miscorrection as capacity and defect risk increase.
Frame splitting with adaptive parity and FEC-aware compression improves real-time recovery from partial burst packet losses in videoconferencing.
Structured ECC syndrome analysis detects adjacent two-bit errors without extra parity bits, avoiding SEC miscorrection and added hardware.
Three-syndrome decoding identifies error boundaries in memory codewords, correcting valid errors while preventing extra corruption beyond symbol limits.
Bounding-box latch grouping and location-aware ECC/BEC interleaving correct multi-bit radiation errors with lower circuit overhead.
A parity-check matrix extends SECDED coding to detect burst errors and reduce silent data corruption in IC data transmission.
Adaptive link training selects interleaved or non-interleaved RS-FEC to balance latency and burst-error resilience in 100 Gb/s Ethernet.
Tests contiguous erasure patterns with RS decoding to correct burst and bit errors in high-speed SerDes transmission.
Adaptive link training selects interleaved or non-interleaved RS-FEC to balance burst-error protection and latency in 100 Gb/s Ethernet links.
Selective bit shifting and parity grouping let DRAM ECC correct single-bit and adjacent two-bit cell errors, improving yield and reliability.
Trapping set core detection and important sampling cut LDPC error floor verification time while enabling reliable low-error code generation.
Tests contiguous erasure patterns in encoded blocks to correct high-speed SerDes burst errors beyond conventional DFE and FEC limits.
Capturing data, check bits, and error syndromes only when needed cuts staging registers, power dissipation, and memory debug uncertainty.
Maps second-type parity bits to bad memory cells so LDPC decoding handles burst errors with less throughput and power impact.
Anchor-point detection and retry re-reads realign a media sector after a missed sync mark, enabling user data recovery with less overhead.
A split parity and ECC layout corrects two consecutive bit errors and detects four while reducing memory area, power, and performance overhead.
Blockwise bit inversion with common-mode signaling cuts burst errors in high-speed serial links without large interleaver memory.