Research And Implementation Of Privacy-Preserving Task Matching In Mobile Crowdsensing

Mobile Crowdsensing is an emerging distributed data collection paradigm.It collaborates through the mobile Internet to realize the distribution of sensory tasks and the collection and use of sensory data in the cloud,so as to use group collaboration to complete large-scale and complex social sensory tasks.Due to the large and growing number of sensing tasks,it is particularly important to provide task matching services for mobile crowdsensing platforms.In the current research,task matching needs to establish the relationship between the perceived user and the task by the task's requirements and user's attributes.However,both task requirements and perceived user attributes are sensitive information.Once the matching information is leaked,the privacy of participants will be seriously threatened.In addition,in the process of participating in the sensing task,the sensing user will have some selfish or malicious behaviors due to additional benefits.These behaviors will directly affect the quality of the sensing data,which will lead to unreliable sensing task results.Therefore,designing a privacy-preserving task matching system with the capability of malicious behavior prevention is a challenge facing the development of mobile crowdsensing applications.Aiming at the above problems,this paper explores a task matching method that is safe and effective and protects privacy in mobile crowdsensing,and has done the following work:1.This article first analyzes the privacy leakage and practical problems in the task matching process,introduces the system architecture and threats of the proposed privacy-preserving task matching scheme,and defines a set of system requirements in detail for the threat model.2.This paper constructs a two-way privacy-preserving task matching scheme,which provides reliable task publishing and accurate task retrieval functions on the premise of not revealing the sensitive information of both users and task.3.This paper considers the possible duplicate data upload behavior and erroneous data upload behavior of perceived users,and proposes a modified group signature scheme to achieve anonymous identity authentication of the perceived users while ensuring the unlinkability of the signature.Through the method of generating binding information of users and tasks,it can effectively prevent selfish repeated uploads and achieve accountability for wrong data sources.4.In order to verify the validity and correctness of the proposed scheme,this paper implements a prototype system,carrying out detailed comparison experiments and performance tests from theoretical analysis and actual cost,and conducts security analysis on the privacy characteristics of the scheme.The experimental results show that the cost of this research scheme has lower overhead,more complete privacy features,and it can protect privacy in real mobile crowdsensing applications.