Research On Privacy Protection In WiFi Fingerprint Indoor Localization System Based On Differential Privacy

With the development of mobile communication technology,users have higher and higher requirements for the accuracy of the real-time indoor location information.Indoor localization based on WiFi Fingerprint is recognized as one of the most promising technologies in the field of indoor localization.The advantages of this technology are obvious,low cost and high reliability.However,the potential privacy issue cannot be ignored.There is a risk of exposing the location information of the individuals who provide WiFi fingerprints during the offline sampling stage.Moreover,the location information of the to-be-localized(TBL)client and the data privacy of the server database may also be revealed during the online localization stage.For the privacy issues in WiFi Fingerprint indoor localization system,the main contributions of the thesis are as follows:(1)To solve the privacy issue in the online localization stage,this thesis firstly introduces differential privacy into the online localization stage of WiFi Fingerprint indoor localization,and designs A Differential Privacy-Based Privacy-Preserving Indoor Localization Mechanism which is named DP3.DP3 is composed of four phases,AP Fuzzification and Location Retrieval in client side,and DP-based Finger Clustering and Finger Permutation in server side.Specifically,in AP Fuzzification phase,instead of providing the measured full finger(including AP sequence and the corresponding RSS),a TBL client only uploads the AP sequence to the server.Then,the localization server utilizes the DP-enabled clustering to build the fingerprints related to the AP sequence into kclusters,permutes these reference points in each cluster with exponential mechanism to mask the real positions of these fingerprints,and sends the modified dataset to TBL client.At client side,Location Retrieval phase estimates the location of the client with KNN algorithm.At last,a real-dataset is used to verify that DP3 can achieve relatively good localization results,while achieving privacy protection.(2)Due to the random selection of the number of clusters kand the initial center points in DP3,the dataset error and localization error are relatively large.Therefore,DP3+ scheme is proposed based on DP3,in which DP-based Finger Clustering phase is improved with Canopy + DP Finger Clustering.Specifically,the following three aspects are improved:(a)the selection of kvalue;(b)the selection of initial center points;(c)the allocation of the privacy budget,assigning different privacy budgets in the two stages of clustering and permutation.Finally,the experimental results show that DP3+ scheme can obtain better dataset utility and localization accuracy than these of DP3 scheme.(3)To solve the privacy issue in the offline sampling stage,differential privacy is applied into the offline sampling stage.A privacy protection scheme based on the crowdsourcing offline sampling stage of WiFi fingerprint indoor localization is designed.Each participant sends the noised measurement data(the noised RSS of all APs)to the service provider instead of sending the original data.The service provider calculates the average signal strength of each AP according to the noised measurement data,obtains the WiFi fingerprint of each location,and finally generates a fingerprint database.In the end,the experimental results show that the scheme can obtain a practical fingerprint database while protecting the privacy of participants' location.