Mobile crowd sensing task allocation method based on Monte Carlo position fingerprints

Abstract

The invention provides a Monte Carlo position fingerprint-based mobile crowd sensing task allocation method, which comprises the following steps of: firstly, determining a corresponding candidate worker set for each task by utilizing the collected fuzzy position information of a to-be-executed task and the fuzzy position information of mobile workers; and then a complete bipartite graph about tasks and workers is constructed on this basis, theoretical optimal matching about the tasks and the workers is further solved by using a KM algorithm, and finally an actual optimal task allocation strategy is obtained by deleting meaningless matching. According to the method, efficient and accurate task allocation can be realized on the premise of protecting crowd sensing participants and bilateral position privacy of tasks.

Description

technical field [0001] The application of the present invention relates to task allocation in crowd sensing, in particular to a mobile crowd sensing task allocation method based on Monte Carlo position fingerprints. Background technique [0002] The ubiquitous portable smart devices and the various sensors on them enable crowd sensing to collect data extensively at a low cost, and a large number of applications based on mobile crowd sensing have begun to emerge. In the crowdsensing system, task assignment is the basis for implementing crowdsensing. The crowdsensing platform needs to assign each task according to the location information of the requested task submitted by the task requester and the current location information uploaded by the mobile worker. ) to many users participating in group intelligence sensing. During the assignment process, it is generally based on specific optimization goals, including minimizing travel distance and maximizing overall revenue. On the...

AI Technical Summary

Problems solved by technology

[0004] Wang introduced the differential geographical obfuscation mechanism into mobile crowd sensing task assignment to protect the location privacy of participants, but this method does not provide location privacy protection for the tasks to be performed. In addition, the real location of mobile workers in this method The operation of fuzzing information leads to inaccurate task assignment results

Yuan designed a grid-based location protection method, specifically using the Chameleon hash function to encrypt the horizontal and vertical coordinates of the center point of the grid where the task or worker is located, but in this method, the task and The exact distance between workers cannot be calculated, and the crowd-sensing platform can only identify mobile workers who are within a certain distance from the task to be performed. Therefore, the task assignment performed by it is also not an accurate assignment

In addition, Ni designed a matrix-based location matching mechanism, so that the crowd-sensing platform can achieve location-based task assignment without disclosing the real location information of mobile users and the perception area of ​​the task, but the matrix-based location Protection does not take into account the exact distance between the worker and the task to be performed, and cannot perform precise task allocation, but can only perform a feasible but rough area-level allocation

Zeng uses multiple fog nodes to protect the privacy of tasks and users' location in the task allocation of mobile crowd sensing by using a multi-secret sharing scheme, and proposes an adaptive top-k worker selection algorithm, so that the crowd sensing platform can serve each waiting group. Execution tasks accurately select the most suitable worker, but this method brings a heavier communication load and computational overhead

[0005] The above methods can ensure the location privacy protection in the task allocation process, but these methods themselves have the disadvantages of inaccurate and lightweight task allocation.

For example, introducing a differential privacy mechanism to probabilistically fuzz the location information of mobile workers will lead to inaccurate user location information obtained by the crowd-sensing platform, which in turn will lead to a decrease in the accuracy of task assignment results. Often brings high communication load and computational overhead

Images