Researches On Privacy Protection Technologies Of Incentive Mechanism And Data Aggregation In Mobile Crowdsensing

As the development of wireless network and the popularity of advanced sensor technologies,various mobile intelligent devices integrating computation,communication,and sensing capabilities are widely applied in our daily life.Many emerging technologies are derived at the same time.Mobile crowdsensing(MCS)is one of the most influential technologies among these emerging technologies.MCS crowdsources a sensing task to a crowd of intelligent mobile devices to collect and analyze sensing data and extracts useful information to complete the task.Compared with traditional wireless sensor networks,MCS has many advantages(e.g.,high flexibility,low cost,strong scalability,wide application range,and rich data)which facilitate the application of MCS in smart city.At present,MCS has been applied to many fields in smart city,such as smart transportation,smart healthcare,environmental monitoring,service recommendation,etc.Although MCS has promising prospects of applications in smart city and future society,there are still many security and privacy challenges in MCS.For instances,challenges include how to protect the privacy of users and requestors,how to design a reasonable incentive mechanism to encourage users to participate in tasks and submit reliable sensing data,how to ensure the integrity of data and the verifiability of task results during data aggregation,and how to efficiently aggregate the data of multiple tasks from multiple requesters.There are a lot of schemes proposed for privacy protection,incentive mechanism design,data reliability,and data aggregation of MCS.However,current solutions exhibit following deficiencies.1)Existing works fail to address the conflicts between privacy protection,data reliability,and incentive mechanism design to achieve a balance among them.2)Few data aggregation schemes solve the privacy protection and data integrity simultaneously while the privacy of task requesters is ignored.3)Scenarios of concurrent multiple aggregation tasks from multiple requesters become increasingly common,but there are few works to solve the contradiction between privacy protection and efficient data aggregation for multiple tasks scenarios.To address aforementioned problems and challenges,this dissertation conducts threefold research on both data collection stage and data aggregation stage of MCS.In data collection stage,to solve the contradiction between privacy protection and evaluation of data reliability when the incentive mechanism is designed,a privacy-preserving and source-reliable incentive mechanism for MCS is proposed in this dissertation,named PRISM.The source-reliability of sensing data is verified.Then the data quality of source-reliable data is evaluated.Finally,the relationship between data reliability and incentive mechanism is established through data quality.By these measures,contradictions among data privacy protection,data reliability and incentive mechanism design are solved.In addition,PRISM designs a new reward distribution method to provide fixed rewards for all users who provide source reliable data and to award extra floating rewards according to the data quality of source reliable data.This reward distribution method makes the incentive mechanism more flexible.A large number of evaluations conducted on real data sets show that PRISM is efficient and provides more rewards for users who collect source-reliable and high-quality data.In the stage of data aggregation,a privacy-preserving and verifiable data aggregation scheme based on fog computing is proposed for MCS in this dissertation(called PriVDA)to protect the privacy of task requesters and to ensure the integrity of transmitted data and the verifiability of task results when protecting the privacy of users.A two-tier aggregation framework based on fog computing is constructed and a phased aggregation method is implemented to reduce the burden of the server.Combining Paillier homomorphic encryption with secret sharing,a privacy protected data aggregation method is designed to protect the privacy of both users and task requesters.PriVDA ensures the integrity of transmitted data and requesters can verify the correctness of task results.In addition,PriVDA can tolerate “bad fog nodes”,i.e.,the existence of some bad fog nodes will not affect the correctness of task results and the reliability of system.Detailed privacy analyses and a large number of experimental results show that PriVDA is privacy-preserving and efficient,and the bad fog nodes has no impact on the reliability and accuracy of the MCS system.In the stage of data aggregation,although PriVDA can solve the problems of privacy protection and data integrity,it can process only one aggregation task at one round which is not suitable for scenarios with multiple concurrent tasks from multiple requesters.To aggregate data of multiple tasks from multiple requesters efficiently while protecting privacy,a fog-assisted privacypreserving and efficient multi-task data aggregation scheme is designed for MCS,named PREMAD.Based on multi-secret sharing and asymmetric bivariate polynomials,a new method is proposed in PREMAD,which can be utilized to perform both data processing and aggregating of multiple tasks securely.While protecting the privacy of users,PREMAD protects the privacy of aggregation result of each task requester as well.Moreover,PREMAD aggregates the data of multiple concurrent tasks at one round.In particular,the privacy-preserving decision vectors are constructed to protect the privacy of users’ decision.In addition,PREMAD can resist the collusion attacks launched by certain entities in the MCS system.Detailed security analysis proves that PREMAD is privacy-preserving and collusion-resistant.Both theoretical analysis and experimental results show that PREMAD is efficient.