Mechanism Design For Time And Location Based Sensing Task Allocation

In the mobile crowdsourcing system,there are many time-and-location-based tasks,such as collecting congestion information for different roads during off-peak hours.Specifically,workers in the mobile crowdsourcing system can collect location-related sensing data to complete these tasks under the constraint of time,with the assistance of smartphones.Considering the cost brought by executing sensing task,the system can assign sensing tasks overlapped in time periods to the same worker with the purpose of reducing the cost of the execution.In order to motivate workers to complete sensing tasks,the system needs to pay rewards to the workers who complete the sensing tasks.In the real world scenario,workers often provide private information strategically to obtain more rewards and maximize their own utilities,such as providing false cost information and working time information.However,the existing algorithms cannot produce efficient allocation strategies when sensing tasks are with overlapped time and workers are untruthful.Therefore,this paper considers how to design mechanisms to motivate workers to provide private information truthfully and maximize social welfare that defined as the total utilities of workers and crowdsourcing platforms.Firstly,the offline mechanism for location homogeneous tasks is studied.Both sensing tasks' and workers' information are reported to the system ahead of time in the scenario.In this section,a VCG mechanism based on dynamic programming optimal task allocation is proposed,and rigorous theoretical analyses proved that the proposed mechanism achieves truthfulness,individual rationality and computational efficiency.And the experimental results showed that the proposed mechanism can get the highest social welfare and guarantee the truthfulness of workers.Secondly,the online mechanism for location homogeneous tasks is studied.In this scenario,the sensing tasks and the workers arrive at the system dynamically,and the system needs to decide to allocate the sensing tasks immediately.An online mechanism based on second price auction is proposed,and rigorous theoretical analyses proved that the proposed mechanism can prevent the workers from reporting unit cost highly.And the experimental results showed that this mechanism outperforms the compared algorithm and can achieve more social welfare.Thirdly,the online mechanism for location heterogeneous tasks is studied.The sensing tasks and the workers have heterogeneous location information.After modeling the location information,a second-price-auction-based online mechanism considering location is proposed.And the experimental results showed that this mechanism outperforms the two compared algorithms and can achieve more social welfare.