Browse Prior Art Database

Interleaved Error Correction for Error Propagation Channel Codes

IP.com Disclosure Number: IPCOM000042849D
Original Publication Date: 1984-Jun-01
Included in the Prior Art Database: 2005-Feb-04
Document File: 2 page(s) / 36K

Publishing Venue

IBM

Related People

Martin, GN: AUTHOR [+2]

Abstract

Arithmetic coding has recently been suggested for the implementation of complex channel codes, especially for high density hard magnetic disks. One problem is that arithmetic codes cause error propagation, which complicates error correction. One way to solve this is to switch the order of application of the channel and error codes; however, this is fairly complicated and loses symmetry of error correction. We propose a method which forces a limit to the error propagation, based on the interleaved error code of the IBM 3950 (Avrind M. Patel, "Error Recovery Scheme for the IBM 3850 Mass Storage System," IBM Journal of Research and Development 24, 32-42 (January 1980). We use exactly the same method as Patel for breaking a segment into sections B0 ... B14, and for writing the coded data to disk.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Interleaved Error Correction for Error Propagation Channel Codes

Arithmetic coding has recently been suggested for the implementation of complex channel codes, especially for high density hard magnetic disks. One problem is that arithmetic codes cause error propagation, which complicates error correction. One way to solve this is to switch the order of application of the channel and error codes; however, this is fairly complicated and loses symmetry of error correction. We propose a method which forces a limit to the error propagation, based on the interleaved error code of the IBM 3950 (Avrind M. Patel, "Error Recovery Scheme for the IBM 3850 Mass Storage System," IBM Journal of Research and Development 24, 32-42 (January 1980). We use exactly the same method as Patel for breaking a segment into sections B0 ... B14, and for writing the coded data to disk. We have several segments per track (rather than per stripe). We use precisely Patel's error code, though an arbitrary channel code. For hard disk applications we probably use longer sections and shorter synchronization patterns than Patel does. The novelty of the method lies in the decode, especially where errors are detected. We buffer all 15 sections prior to channel decoding, and apply channel decode on them in parallel to produce 15 source bytes (B14-0...B0-0).

We then apply the error correction code to this first codeword. If errors are found, we recode the byte for the sections in error. This recoding puts the channel coder to a given state at the end of the byte. We use this state to reset the end of byte deco...