Research on Task Distribution Mechanism Based on Public Transit System in Crowd Sensing
Jian AN,Zhen-Long PENG,Xiao-Lin GUI,Le-Le XIANG,Dan-Wei LIANG
DOI: https://doi.org/10.11897/SP.J.1016.2019.00295
2019-01-01
Abstract:As mobile smart devices become increasingly popular and are equipped with increasingly powerful sensors, they have been pervasively applied in crowdsensing as effective tools to solve large-scale sensing tasks in urban areas.Task requesters can allocate sensing tasks to mobile nodes through a crowdsensing platform, eliminating the cost of deploying and maintaining large numbers of fixed sensors.However, some crowdsourcing tasks (e.g.audio and video transmission) that generate a large amount of sensed data may bring higher network traffic costs to participants using a 3 G/4 G network, which may affect their enthusiasm and cause enormous transmission pressure on the network.To reduce communication costs, if sensed data does not require realtime uploading, researchers have applied opportunistic network and "store-carry-forwarding"totransmit data.But they have ignored a significant problem that it is impossible for a large amount of data to be transferred in such a short momentary encounter.With the increasing sensing capabilities of smart devices, the types and sizes of sensed data have changed greatly.The data transmission paradigm in an opportunistic network is based on the assumption that data exchange can be completed as long as the mobile nodes can connect with each other.It implies that the data size can be ignored or that the data can be exchanged in a transient encounter between mobile nodes.In this paper, we built a data diffusing and transmission paradigm in crowdsensing based on a City Public Traffic System (PTS), and emphatically thoroughly discuss a paradigm for Multi-Task diffusion and transmission within budget constraints.This paradigm makes full use of the advantages of a bus to realize the rapid transmission of large-scale sensed data:predictable trajectory, wide coverage area, fast moving speed and long contact duration among passengers.First, we use the division of voronoi diagram to divide the target area and achieve the reasonable coverage of sensing tasks and target area.Secondly, we analyze the problem of task differential distribution under certain budget constraints, and propose two distribution algorithms:distribution algorithm based on coverage difference (COV-DA) and distribution algorithm based on diffusion difference (SPR-DA).Both transmission schemes are under budget constraint that aims to maximize the overall transmission utility of all data brought into the PTS.By diffusing and scheduling all data in the PTS, the data can be uploaded to the crowdsensing platform by proper mobile passengers.Finally, to effectively evaluate the performance of our algorithm, the COV-DA and SPR-DA algorithms are compared with the classical Greedy and Epidemic algorithms in detail from the accuracy of task distribution, the average moving distance and the time of task distribution.The experimental results show that the SPR-DA algorithm has better distribution accuracy, while using the COV-DA algorithm, the mobile node needs a shorter moving distance when the task is completed.Compared with Greedy and Epidemic algorithm, COV-DA and SPR-DA have better performance in distributing accuracy and average mobile distance, and can achieve a tradeoff between overall transmission utility and transmission redundancy, and save more network traffic costs and other resources for mobile nodes.