A Fragile Watermarking Scheme For Tamper Detection Of Mutidimensional Database

With the rapid development of computer technology,data is increasing faster and faster in all fields.It seems difficult that multidimensional database technology such as traditional serial data warehouse to adapt to the rapid development of application requirements.Multidimensional database is a multidimensional matrix,where users can analysis and dive deeply into with different dimension and level.It has been used widely in OLAP which is one of most popular decision support technology in the market recently.In various safety issues,the information attack of multidimensional database has been one of the most damaging threaten in e-commerce business and Internet system.Scientific researchers are studying actively in those related safety topics.This argument provides a constructive approach of multidimensional database model with a watermarking scheme for tamper detection of multidimensional database.We separate data information unit into independent groups,then strategically rank the order of each group's data information unit by rank algorithm,and put the authentication into structured multidimensional database.The whole embedded process of fragile watermark will neither take redundant information nor affect the data's regular usage.This method can verify whether the carrier of data information was modified or not by validating the watermark damaged state.With higher non-perception to user,the modified information group can be located.This method takes short time in authentication embedding and testing which is easy and feasible.We built the Oracle cube and finish a large number of simulation experiments based on the tamper detection algorithm of multidimensional database to demonstrate the validity of the algorithm.We also complete performance analysis of the algorithm with a large number of simulations to verify the applicability and feasibility of our scheme.A large number of simulation experiments show that our algorithm is simple and easy to realize.The detecting process costs less time,the performance is stable and has strong practicability in real applications.