Browse Prior Art Database

Data Security and Recovery Technique

IP.com Disclosure Number: IPCOM000076675D
Original Publication Date: 1972-Apr-01
Included in the Prior Art Database: 2005-Feb-24
Document File: 2 page(s) / 14K

Publishing Venue

IBM

Related People

Goldberg, SL: AUTHOR [+2]

Abstract

Most existing error recovery techniques are designed to correct a single bit or byte. On a disk, it is more likely that an entire track will be lost. Described is a method for storing data which permits recovery of all data on a track or other size physical record. In addition, the storage format is such that it may be used as security against unauthorized access.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 2

Data Security and Recovery Technique

Most existing error recovery techniques are designed to correct a single bit or byte. On a disk, it is more likely that an entire track will be lost. Described is a method for storing data which permits recovery of all data on a track or other size physical record. In addition, the storage format is such that it may be used as security against unauthorized access.

This method uses k+1 tracks to store k tracks of data. It writes the tracks T(0), T(1), ..., T(k) in the following sequence: T(0), (T(0)VT(1)) (T(1)VT(2)), ... (T(k-1)VT(k)), T(k). Each track, except the first and last, is written as the exclusive ORed (V) result of it and its preceding track. If any one of the k tracks of data are lost, that data can be recovered by reading the remaining tracks, and exclusive ORing each track with the next track. The result is the recovered track of data.

Data in this format can easily be used when processed as a sequential file. It requires the use of two track length buffers in the following manner: The first track (T(0)) is read into buffer 0. The second track (T(0)VT(1)) is read into buffer 1 and exclusive ORed with the contents of buffer 0, giving the data in T(1). Now the next track is read into buffer 0 and exclusive ORed with buffer 1. This process is repeated until all data has been processed.

Should an error occur while processing this information, recovery is possible by the following method: Suppose the track i is lost. (This contains data (T(i- 1)VT(i))). One starts by reading in the last track (T(k)) onto buffer 0. The second buffer (buffer 1) will now be filled with the preceding data (T(k-1)VT(k)) permitting one to recover the data T(k-1) by exclusive ORing the two buffers. This process is repeated until one recovers the data T(i).

Example: Suppose the file consists of five tracks of data, A, B, C, D, and E. Using this technique, the data would be written out as six tracks containing the following information: A, AVB, BVC, CVD, DVE, E.

To read the tracks in sequence, one must do the following: Read in the track containing A. Read in the track containing AVB and exclusive OR this with A, giving B. To get C, read in the next track (BVC) and exclusive OR this with B. This process is continued to get D and E.

Suppose that the track containing CVD is lost. Both C and D can be obtained by the following process: A can be obtained directly.

B is obtained by exclusive ORing A with the track containing AVB. Now C can be obtained by exclusive ORing BVC with B. To obtain the information contained in D, one starts with the last track, containing E. This is then e...