Research On Privacy-preserving Data Aggregation Mechanism Based On Local Differential Privacy In Smart Grid

Smart grid can provide more intelligent services than traditional grid by introduc-ing advanced information and communication technology(ICT)in the grid.In smart grid,smart meters installed at user side report the electricity consumption data to the control center periodically.Through the data collection,aggregation and analysis of smart grid,the control center can estimate the real-time power consumption of the grid,so as to manage the generation,transmission and distribution in smart grid.Further-more,the control center can make dynamic electricity price strategy according to the sta-tistical results.Although the development of smart grid has a bright future,the privacy protection problem must be solved in the process of smart grid development and popu-larization.Data reported by smart meters contain user’s private information.Through the analysis of user’s power consumption data,a user’s power consumption mode can be obtained,which poses a great threat to his/her privacy.For example,a user’s daily activity pattern can be easily inferred from the power consumption pattern,and thus de-termine whether he/she is at home or not.Therefore,it is necessary to propose efficient privacy preserving data aggregation scheme in smart grid.In smart grid,the data acquisition task is periodic and frequent.Smart meters need to submit real-time power consumption data to the control center every other period of time(for example,15 minutes).In addition,smart meters installed in households have limited computing resources.This requires that the computing cost of smart meters in the aggregation process should be reduced as much as possible while meeting the requirements of privacy protection.At present,the common privacy protection data aggregation schemes in smart grid either have higher requirements for the computation.and communication ability of smart meters,or need a trusted-third-party to coordinate in the system.Aiming at these problems in practical application,this paper proposes practical privacy protection data aggregation schemes in smart grid.Moreover,simple collection of users’ total power consumption data can not meet the demand of smart grid for power grid data analysis.In order to provide more intel-ligent services based on power grid data analysis,the data that need to be collected are not only the users’ total power consumption data,but also the fine-grained collection of users’ data.It is more necessary to collect user data in a fine-grained way,so as to ana-lyze the operation of power grid and judge the power demand in a more comprehensive and detailed way.In this process,the privacy protection of users’ data becomes more important.Therefore,we need to consider more fine-grained privacy protection data aggregation in smart grid,so that smart grid can achieve more privacy protection data analyses.This paper proposes corresponding privacy protection data aggregation schemes for the privacy protection of users’ data in smart grid data aggregation scenarios.Specif-ically,the main contributions of this article are as follows:·Considering the demand of frequent data collection in smart grid and the limited computing capability of smart meters,this paper proposes an efficient privacy-preserving data aggregation scheme in smart grid.In this paper,a lightweight pri-vacy preserving data aggregation scheme in smart grid is proposed by introducing Local Differential Privacy(LDP)framework into smart grid data aggregation sce-nario.In this scheme,users disturb the data generated by the smart meter through local randomized response without trusted third party.Meanwhile,the scheme can effectively support the dynamic joining and quitting of users without vast ad-ditional computation or communication overhead.In addition,we consider more special scenarios in smart grid data aggregation,and propose a grouping based scheme for large-scale data aggregation,which enables our scheme to deal with most scenarios in smart grid.We simulate the implementation of our proposed scheme,and analyze the performance and security.The analysis shows that the proposed scheme has less computation and communication overhead while en-suring the utility and privacy protection.·In addition to simple data aggregation,in order to achieve more detailed and com-prehensive data analysis in smart grid,it is necessary to classify and fine-grained collection of user power consumption data.In this process,the users’ data pri-vacy also need to be protected.Therefore,this paper designs a classified privacy-preserving data acquisition scheme in smart grid which satisfies local differential privacy.Users classify and upload the data generated by smart meters.The con-trol center can provide more precise and intelligent services for smart grid by ana-lyzing different types of power consumption data.Based on the local differential privacy framework,we design a data aggregation scheme to protect classified data privacy.We simulate the proposed scheme,and analyze the performance and security.The analysis shows that the proposed scheme can provide more de-tailed data analysis in smart grid under the premise of protecting the privacy of user data.Meanwhile,the computation and communication costs are acceptable.In conclusion,by introducing local differential privacy framework into smart grid data aggregation scenario,an efficient privacy preserving smart grid data aggregation scheme is realized.Furthermore,considering the data analysis requirements in smart grid,a privacy-preserving classified smart grid data aggregation scheme is proposed.In addition,this paper analyzes the security and performance of each scheme in detail to further verify the security and practicability of the scheme.