Research On Data Hidden Based On InnoDB

Data hiding is one of the effective techniques for solving data security and data attacks.As a place for data storage and protection,the database storage engine needs to guarantee the security of the stored data.The database widely uses technologies such as logs and indexes to improve storage performance,which also brings spatial redundancy to the storage of the database,thereby providing the possibility of data hiding.At present,data hiding research based on databases is mainly limited to small-capacity data hiding in non-relational databases.Therefore,in-depth research on efficient data hiding methods for different levels of data in Inno DB is of great significance for improving data security in the database.On the basis of comprehensive analysis and research on the index and deletion mechanism in Inno DB,firstly,low-level data hiding is carried out in Inno DB.Aiming at the index and deletion mechanism and characteristics in Inno DB,a method of using Inno DB index structure and deleting list pointers is proposed: by removing the target from secondary index tree,so that the related records cannot be accessed by the query statement using the secondary index;by reorganizing the data page directory and disconnecting the corresponding primary index,the record is invisible to the query of the primary index;by changing the delete operation,the deleted records are not located at the deleted list,making the relevant data completely invisible.Then,in order to carry out higher-level data hiding in Inno DB,only using the index and delete mechanism will bring great performance overhead.By analyzing the structure of Inno DB data pages,a data hiding method based on the idle space of Inno DB compressed data pages is proposed.The hidden data is compressed and put into the user record space,and other spaces remain unchanged.Its reliability is ensured through the transfer mechanism.The experimental results show that for low-level data hiding in Inno DB,compared with no data hiding,this method is feasible and only reduces the read and write performance by 1%under optimal conditions.Compared with low-level data hiding methods in Inno DB,higher-level data hiding can increase read and write performance by 10%.