Location Privacy Protection Methods In Task Allocation For Mobile Crowdsensing Over Road Networks

Mobile Crowdsensing(MCS)is an efficient sensing paradigm that enables users to collect sensing data using mobile devices.Many MCS platforms require users to provide real-time locations for task allocation.However,the sharing of location will bring serious privacy leakage problems.Some users refuse to participate in tasks due to privacy concerns,resulting in a lower task completion rate.To solve the problem of users’ location privacy in task allocation for MCS,both academia and industry are actively exploring effective technical means.Differential privacy is a mainstream protection method adopted in relevant research.By providing the location after differential privacy processing to the platform,the platform can obtain the obfuscated location of the user instead of the real location,to protect user’s location privacy.Although some methods based on differential privacy have been proposed,they do not consider the influence of road networks and continuous sensing on privacy in MCS.Therefore,this thesis focuses on location privacy in task allocation for MCS over road networks.The main work is as follows:(1)To protect users’ location privacy in task allocation for MCS over road networks,we propose a privacy protection task allocation framework.The framework is divided into two parts: 1)To obfuscate users’ locations,we propose a protection method based on GeoGraph-Indistinguishability(Geo GI);2)An optimal task allocation scheme based on obfuscated locations is proposed to minimize the travel distance of users to complete all tasks.Experimental results show that the proposed framework is superior to the existing differential privacy protection methods in terms of privacy and task allocation efficiency.(2)To solve the problem of privacy leakage caused by continuous locations in task allocation for MCS,we propose a privacy protection method based on -Differential Privacy.Firstly,a first-order Markov model is used in modeling users’ mobility.Secondly,analyze and quantify the privacy leakage caused by an adversary.Finally,the differential privacy method is improved through the privacy budget allocation strategy.The experimental results show that the improved method can resist the attack well and has a good performance in terms of task allocation efficiency.