The rapid adoption of the vehicles and their on-board sensors as a primary means of transportation make them natural candidates for the outsourcing of data collection. However, vehicles mobility patterns tend to cluster into specific regions such as highways and popular roads, that makes their utilization difficult for data collection in isolated regions with low density traffic. We tackle this problem by proposing a probabilistic incentive mechanism for Vehicular Crowdsensing (VCS) that encourages vehicles to deviate from their pre-planned trajectories in order to visit and collect data from the isolated places. Our proposed framework is able to handle asynchronous vehicles. Also, vehicles consider the traffic holistically to find more profitable routes. By using a realistic vehicular movement data set (UBER movement), open-street maps (OSM) and SUMO vehicular traffic simulator, we show our algorithm significantly outperforms traditional approaches for trajectory generation in terms of spatial and temporal coverage, road utilization, and average participant utility.

中文翻译：

---

用于 AV 的车辆众感市场

车辆及其车载传感器作为主要交通工具的迅速采用使它们成为数据收集外包的自然候选者。然而，车辆移动模式倾向于聚集在特定区域，例如高速公路和热门道路，这使得它们难以在交通密度低的孤立区域收集数据。我们通过为车辆人群感应 (VCS) 提出一种概率激励机制来解决这个问题，该机制鼓励车辆偏离其预先计划的轨迹，以便访问和收集偏远地区的数据。我们提出的框架能够处理异步车辆。此外，车辆会从整体上考虑交通情况，以找到更有利可图的路线。通过使用真实的车辆运动数据集（UBER 运动），