Research On Privacy Protection Of Multiple Cloud Service Requests Based On Risk Assessment

The development of cloud computing comes with security and privacy issues.Traditional researches of cloud security mainly focuse on the security and privacy issues of cloud data and user,for the privacy issue of cloud service requests they have less involved,and at the same time their solutions are not suitable for this problem.The existing effective method is noise obfuscation,but the current researches are all for solving the problem of privacy protection of single cloud service request within a certain time-slice,for the scene of multiple cloud service requests that occur at once their research results are not that effective and suitable.This thesis studies how to control the noise cost while ensuring the security of privacy information in the scene of multiple cloud service requests privacy protection.The main works are as follows.Stating the reasons why existing methods could not meet the needs of noise cost control and safety in the scene this thesis proposes b y analyzing the research status of single cloud service request privacy protection in cloud computing environment.According to the features of multiple cloud service requests privacy protection mentioned in this thesis,introducing the concept of risk assessment and proposing important-degree and joint derivation probability of sensitive information in cloud service requests as the two risk parameters.A risk assessment algorithm based on D-S evidence theory is proposed.This algorithm models the relationships of sensitive information and uses the multi-source evidence fusion rules of D-S theory to complete the quantization of the two risk parameters,and then construct the expression for ri sk calculation.Improving the original strategy of noise generation and obfuscation and using the improved strategy to optimize the mechanism of multiple cloud service requests privacy protection according to the results of risk assessment.A risk assessment algorithm based on fuzzy theory is presented,it uses the definition of membership degree in fuzzy theory and fuzzy comprehensive assessment to accomplish the quantization of the risk parameters,and the same construct an appropriate risk calculation expression.Improving again the noise generation and obfuscation strategy,and then using it to continue optimizing the mechanism of multiple cloud service requests privacy protection.Through the comparison of the simulation results with the existing method,the advantages and effectiveness of noise cost control and safety of the two privacy protection methods based on different theories that are proposed in this thesis are being proofed.Finally this thesis analyzing the similarities and differences between the two methods,summarizing their suitability for different scenarios of multiple cloud service requests privacy protection.