Browse Prior Art Database

Magneto-Optic Data Integrity Using CRC Checksum

IP.com Disclosure Number: IPCOM000111543D
Original Publication Date: 1994-Mar-01
Included in the Prior Art Database: 2005-Mar-26
Document File: 4 page(s) / 171K

Publishing Venue

IBM

Related People

Klem, RA: AUTHOR [+4]

Abstract

The data integrity requirement for data transmitted between the host and an optical disk recording system is to have less than one detected data error in 10 to the 17 bytes transferred. As greater amounts of data are stored and retrieved from optical systems, this requirement will increase. The following describes a means of improving the data integrity of optical disk recording systems to meet the increased requirement.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 35% of the total text.

Magneto-Optic Data Integrity Using CRC Checksum

      The data integrity requirement for data transmitted between the
host and an optical disk recording system is to have less than one
detected data error in 10 to the 17 bytes transferred.  As greater
amounts of data are stored and retrieved from optical systems, this
requirement will increase.  The following describes a means of
improving the data integrity of optical disk recording systems to
meet the increased requirement.

      The host transmits/receives blocks of data.  A block of data is
comprised of one or more records, where a record is the fundamental
unit of data transferred between the host and drive.  In the ISO
Standard for magneto-optic recording, this is either 512- or 1024-
byte records and is determined by the media format.  This record is
contained in a sector, which, in addition to data, has control
fields, pointer fields, Reed Solomon Error Correction Code (ECC)
information and CRC which verifies ECC corrections.  This sector
image is recorded on the media.

      The method of data corruption detection will be to append "n"
bytes of CRC/Checksum to the end of a block and "k" bytes of
CRC/Checksum to the end of a record.  The details of this method are
explained below.  Note that the controller is assumed to be contained
within the drive in the following description and the words drive and
controller are interchangeable.  The Figure shows the controller
portion including the CRC/Checksum generator/checker module (GC),
which connects to the buffer bus and to the SCSI interface module.
This permits the GC module to capture the bus from the SCSI interface
module and ODC (Optical Disk Controller) when necessary.

      The scenario for transferring data follows.  Before the host
requests a write operation, it sends a Mode Select command to tell
the drive to expect "n" bytes of CRC/Checksum appended to each block
and a Mode Select command to append "k" bytes of CRC/Checksum data to
the record contained in the sector.  (Note:  This could be done with
the same Mode Select command.)  It is important to have Mode Select
commands for the two modes of operation to maintain compatibility
between various configurations.  The host then requests a write
operation which informs the drive as to the size of the block to
expect.  The host then transmits the block of data and appends "n"
bytes of CRC/Checksum to the end of the block.  In response to the
write request, the drive microprocessor programs the GC module to
expect "n" bytes at the end of the block and that it is a write
operation.  The SCSI interface module is set to expect the number of
bytes contained in the block size, plus "n" bytes of CRC/Checksum.
It then programs the Optical Data Controller (ODC) to expect the
number of records transmitted in the block, where the record size is
the transmitted record size plus "k" bytes of CRC/Checksum.

      The ODC controls the transfer of data from...