Browse Prior Art Database

Error correction and the coding in disk storage device driver.

IP.com Disclosure Number: IPCOM000014835D
Original Publication Date: 1999-Oct-01
Included in the Prior Art Database: 2003-Jun-20
Document File: 2 page(s) / 28K

Publishing Venue

IBM

Related People

Hiroyuki Ono: AUTHOR [+3]

Abstract

Disclosed is a data recovery method for disk storage devices by applying matrix ECC (Error Correction Code) for a block of sectors. This method allows long burst data error correction capability with little loss of the data capacity and the performance. The configuration of a block is, j (integer) sectors are appended as the matrix ECC for leading k (integer) data sectors such as, Sector_1, Sector_2,,,, Sector_k, ECC_sector_1, ECC_sector_2,,,ECC_sector_j

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 2

Error correction and the coding in disk storage device driver.

    Disclosed is a data recovery method for disk storage devices by applying
matrix ECC (Error Correction Code) for a block of sectors.

    This method allows long burst data error correction capability with little
loss of the data capacity and the performance.

The configuration of a block is, j (integer) sectors are appended as the
matrix ECC for leading k (integer) data sectors such as,

      Sector_1, Sector_2,,,, Sector_k, ECC_sector_1,
ECC_sector_2,,,ECC_sector_j

Other wise, the ECC sectors can lead prior to the data sectors.
Each sector is divided into n (integer) symbols such as follows.

S11 S12 S13.... S1n for Sector_1
S21 S22 S23.... S2n for Sector_2

:

Sk1 Sk2 Sk3.... Skn for Sector_k

C11 C12 C13.... C1n for ECC_sector_1
C21 C22 C23.... C2n for ECC_sector_2

:

Cj1 Cj2 Cj3.... Cjn for ECC_sector_j

Where, S and C are the unit of the symbols for the data and the ECC each. Both
symbols have the same length of L (bits).

The set of C1l C2l... Cjl (l = 1,2,...n) is generated for the column of data
set of S1l S2l... Skl.

And the one sector length is (n * L)/8 bytes. Here, k is less than or equal to
2^L - j and j is grater than or equal to 2.

Small error can be conventionally corrected in the sector internally as row
direction of the matrix. And this method allows long burst error correction by
calculating the column sets of the data and ECC symbols. The error sectors are
known as the row data collection prior to the column data...