Abstract
Wireless power transfer (WPT) is a promising technology to prolong the lifetime of the sensors and communication devices, i.e., workers, in completing crowdsourcing tasks by providing continuous and cost-effective energy supplies. In this paper, we propose a wireless powered spatial crowdsourcing framework which consists of two mutually dependent phases: task allocation phase and data crowdsourcing phase. In the task allocation phase, we propose a Stackelberg game based mechanism for the spatial crowdsourcing platform to efficiently allocate spatial tasks and wireless charging power to each worker. In the data crowdsourcing phase, the workers may have an incentive to misreport its real working location to improve its utility, which causes adverse effects to the spatial crowdsourcing platform. To address this issue, we present three strategyproof deployment mechanisms for the spatial crowdsourcing platform to place a mobile base station, e.g., vehicle or robot, which is responsible for transferring the wireless power and collecting the crowdsourced data. As the benchmark, we first apply the classical median mechanism and evaluate its worst-case performance. Then, we design a conventional strategyproof deployment mechanism to improve the expected utility of the spatial crowdsourcing platform under the condition that the workers' locations follow a known geographical distribution. For a more general case with only the historical location data available, we propose a deep learning based strategyproof deployment mechanism to maximize the spatial crowdsourcing platform's utility. Extensive experimental results based on synthetic and real-world datasets reveal the effectiveness of the proposed framework in allocating tasks and charging power to workers while avoiding the dishonest worker's manipulation.
Introduction
Spatial Crowdsourcing (SC) leverages mobile workers to perform location-specific tasks, such as environmental sensing or traffic monitoring. However, two critical challenges often limit its effectiveness: energy constraints of mobile devices and the potential for strategic dishonesty among workers. In remote or emergency zones, workers may lack sufficient battery to complete tasks. Furthermore, when a mobile base station (BS) is deployed to provide power and collect data, workers may selfishly misreport their locations to attract the BS closer to them, thereby gaining more power at the expense of the platform's overall utility.
The Two-Phase Framework
To address these challenges, the authors propose a framework consisting of two mutually dependent phases: task allocation and data crowdsourcing.
- Task Allocation Phase: Modeled as a Stackelberg game, the platform (leader) announces a total power supply, and workers (followers) determine their data transmission rates. This phase ensures an optimal distribution of power based on the workers' capabilities.
- Data Crowdsourcing Phase: The platform deploys a mobile BS to transfer power and collect data. This phase focuses on the facility location problem, ensuring the BS is placed optimally relative to worker locations.
Addressing Strategic Dishonesty
A core contribution of this work is ensuring that workers report their true locations. If a worker misreports their position, the platform's deployment of the mobile BS becomes suboptimal, degrading the system's overall efficiency.
The authors evaluate three mechanisms for BS deployment:
- Median Mechanism (MED): A baseline strategyproof approach using the median of reported coordinates.
- Median Single-Constant (MSC): An enhanced version for known worker distributions.
- Deep Learning Mechanism (MDL): A novel neural network approach that enforces monotonicity to guarantee strategyproofness. By following Moulin’s generalized median rule, the MDL mechanism learns optimal deployment strategies from historical data while remaining immune to manipulation.
Evaluation and Key Results
The mechanisms were tested using both synthetic data and a real-world dataset from the NYC MTA, featuring over 2 million GPS traces.
- Efficiency Gains: The MDL mechanism significantly outperformed the MED baseline, achieving an 18.39% reduction in the worst-case performance ratio in real-world simulations.
- Environmental Robustness: MDL demonstrated high stability even in high path-loss environments where the MED mechanism's efficiency dropped sharply.
- Energy as an Incentive: By using Wireless Power Transfer (WPT) as the primary reward, the system creates a direct physical link between a worker's location and their utility, effectively using energy as a currency.
Conclusion
This research demonstrates that deep learning can be used to "discover" optimal, strategyproof mechanisms for spatial crowdsourcing. By baking truthfulness into the neural network's architecture, the MDL mechanism ensures system efficiency without requiring manual derivation of complex rules. Future work will explore dynamic scenarios with moving workers and multi-facility coordination.
Summary updated