Error correction methodology based on redundant recovery code for optical storage system
Zheng Mu, Luo Tiewei     
Amethystum Storage Technology Co., Ltd., Meizhou, Guangdong 514779, China

Overview: Amethystum Storage Technology Co., Ltd. is a worldwide leading provider of data storage hardware, software systems and solutions. Amethystum host the development of the intelligent storage series products, it supports multiple storage medias including SSD, HDD, Magnetic tape, optical and so on, and offers the optimal solutions for data lifecycle management (DLM), implements long term storage, data security and energy-efficient. In the era of Big Data, Amethystum optical libraries is designed for cold data storage. To improve the data reliability furthermore, Amethystum make a new requirement at software system level.

The redundant recovery code (RRC) is a kind of code that needs more symbols or signals than the number of symbols or signals required to represent the information. It utilizes the coding theory of error correction code which can find and correct the data error automatically during data storage or transmission. The redundant recovery code has a wide applications in the field of information technology. Normally, Hamming code, CRC code and BCH code are popular in the field of information technology.

With the development of optical disc design and manufacture technology, a multiple layers recoding technique has been introduced, it increase data density enormously. However, this technology causes more error data in the same damage on the optical disc. In optical recording roughly two types of errors can be distinguished: single errors and bust errors. Single errors only affect one or two bytes, burst errors are caused by defects on the disc surface like scratches, dust, fingerprints etc. Burst errors recovery is the point in this study. Blu-ray disc physical format specifications adopts an Error-Correcting Code (ECC) and a Modulation Code. They can recover tangent direction burst error with 4.5 millimeter burst error length. In order to improve the reliability of optical storage data, this paper proposes an error correction methodology in optical storage system which based on redundant recovery code technology, it relates to the field of optical disc data storage.

The methodology consists of two opposite processes – recording and retrieving. While recording data, firstly, splitting user data into blocks and encoding it with redundancy recovery code; next, organizing data blocks as UDF (Universal Disc Format) file system; finally, modulating and encoding UDF file system data as strip group according to the optical disc physical format and recording it into optical disc. In contrast, while retrieving data, demodulating and decoding data strip group which is stored in optical disc at first; later, following UDF file system format to resume user data; in the end, verifying user data with redundant recovery code and return it to user. This methodology separates user data as 992 megabytes blocks as 31×32 megabytes length data blocks, then calculates these data blocks with XOR algorithm to get 1×32 megabytes data block, finally, gets 32×32 megabytes mixed data will be recorded into optical disc.

The methodology is compatible with the standard optical disc file system, and improves the fault tolerance efficiently. In the tangent direction, it can recover the whole circumference burst error, the original blu-ray disc standard algorithm has just ability to recover 4.5 millimeter length burst error; in the radial direction, the redundant recovery code is able to recover 33 micrometer length burst error, but the blu-ray standard can do nothing.