Method and apparatus for erasure decoding an ECC coded bitstream
The error correction capability of block codes can be doubled if error locations are known. Prior art approaches for error location detection always involve adding dedicated redundant data which then are evaluated to yield error location information. The present invention proposes and describes how error location information in the form of clues is derived from given DC control bits that are anyway present in a data stream.
1. A method for erasure decoding an ECC coded bitstream that comprises Data Channel (DC) control bit patterns formed from DC control bits of the ECC coded bitstream together with preceding and subsequent data bits from the ECC coded bitstream, comprising:
using the DC control bit patterns as self-checkable erasure clues,
wherein no other information is needed other than the erasure clue itself to derive if the erasure clue has been corrupted or not,
wherein the erasure clues constitute an erasure information where negative erasure clues indicate no error, and positive erasure clues indicate the occurrence of an error.
2. The method according to claim 1 , wherein the erasure information from the bitstream is postprocessed by the following steps:
ignoring isolated positive clues,
connecting non-isolated positive clues,
expanding sequences of positive clues,
flagging as erasures the expanded sequences of positive clues and any channel data blocks that are flanked on at least one side by positive clues.
3. The method according to claim 1 , wherein the erasure information is consolidated with a second erasure information before being used in the erasure decoding.
4. The method according to claim 3 , wherein the consolidating is performed by majority decision among corresponding bits of the erasure information and of the second erasure information.