Collaborative Algorithm And Application Of Matrix In Networked Information System

Since the rise of Internet technology,information technology has developed rapidly,and the ensuing information security has become a research hotspot in information-driven systems.Information security includes anonymity,concealment,and interference.In an informationdriven system,unauthorized users(intruders)should not obtain or even destroy any secret information in the system.The non-intrusive proposal is to prevent such information leakage and avoid intruders from attaining some secret information.Most of the existing methods for judging information security use integer linear programming algorithms and global diagnostic methods,which are complex and difficult to test.In the global diagnostic method,the number of formulas of integer linear programming algorithms will vary as the number of low-level transitions increases.Thus,the fault tolerance is low.In view of the problems in the prior work,first,this thesis proposes a distributed decomposition algorithm to decompose a large complex system into several simple small systems,and verify the consistency of the attributes and states of the system before and after the decomposition,via analyzing the security of the small systems,so as to obtain the security of the original complex system.Secondly,the complex integer linear programming algorithm is converted into a clear and easy-to-understand matrix algorithm,and the security of the system is analyzed by the results obtained by solving the matrix equations.The matrix algorithm displays each node in the system in a matrix,and the operation process is clear and straightforward.When the changes in the system are dynamically triggered,we only need to make changes in the corresponding matrix without changing the overall algorithm.In addition,when the number of low-level transitions in the system increases or decreases,we do not need to increase the number of matrices.Instead,we only increase the matrix dimension.Thus the method is tolerant to fault and flexible in application.Then,the distributed decomposition algorithm is combined with the matrix algorithm,so that the distributed matrix algorithm is proposed.Through the matrix algorithm,the small lowerlevel system is judged and the security of the original upper-level system is obtained.This thesis also introduces and models the colored net,and diagnoses the security of the system through the distributed matrix algorithm,so as to verify the feasibility and effectiveness of the matrix algorithm.Finally,the security detection of the distributed matrix algorithm is performed on typical network cases,and compared with the existing methods.All results fully demonstrate the good performance of the matrix algorithm.The method proposed in this thesis has the following advantages:(1)The distributed matrix algorithm is simple to calculate,tolerate to the fault,and convenient for maintenance.This can compensate the shortcomings of the integer linear programming algorithm that is difficult to calculate for the analysis of complex systems.(2)The distributed matrix algorithm overcomes the shortcomings of the global safety diagnosis method,and divides the index decomposition of the complex system in detail,and decomposes the system according to the index decomposition algorithm,which improves the computational efficiency.(3)Distributed matrix algorithm has obvious advantages in the application of colored nets.The input and output in ordinary nets are all expressed in the form of vectors.Whereas in the colored nets,they are all expressed in the form of matrices.Therefore,compared to integer linear programming algorithm,matrix algorithm is more suitable for judging the security of the colored Petri nets.