Research On Secure Task Assignment In Spatial Crowdsourcing

With the rapid development of mobile devices and mobile Internet,spatial crowdsourc-ing has gained extensive research and industry attention as a new tool of problem-solving.Spatial crowdsourcing is mainly used to complete some location-based tasks.In the pro-cess of spatial crowdsourcing,the workers need to complete the tasks at the fixed location and also need to upload their own private information to a SC platform for task assignment.However,current mainstream researches focus on the quality and efficiency of task assign-ment,ignoring privacy disclosure problem caused by the untrustworthy third parties and the malicious external attackers.Therefore,we first study the travel-time-first and secure task assignment,and further study the skill-matching and secure task assignment.In the research of travel-time-first and secure task assignment,to achieve mutual priva-cy preserving for workers and tasks,we design a secure task assignment protocol based on partial homomorphic encryption.In the task assignment strategy,we take a travel-time-first assignment strategy instead of the traditional distance-first assignment strategy.However,the travel-time-first allocation strategy brings the problem of ciphertext-based division oper-ations.To this end,we propose a safe and efficient algorithm to avoid the division operation on ciphertext.In terms of the performance of the protocol,we also design a parallel com-puting strategy to reduce the large computational overhead due to homomorphic encryption.Finally,we prove the security of the protocol theoretically,analyze the performance of the protocol,and verify the feasibility and the acceptable computing cost of the protocol through comparative experiments.In the research of skill-matching and secure task assignment,to better adapt to the spatial crowdsourcing,we bring the skill attributes of workers and tasks.In terms of privacy preserving,we expand the travel-time-first and secure task assignment protocol to protect the skill privacy of workers and tasks.Based on this secure task assignment protocol,we introduce a task assignment problem for multi-task and multi-workers,and prove that the problem is an NP-hard problem.Then we propose a greedy-based task assignment algorithm and analyze its complexity.Finally,we analyze the security and performance of the protocol theoretically,and verify the acceptable computing cost of the protocol and the effectiveness of the algorithm through comparison experiments.