A stability-based group recruitment system for continuous mobile crowd sensing

摘要

With the proliferation of Mobile Crowd Sensing (MCS), many domain applications that answer different sensing requests, have been benefiting from the availability of participants in areas of interest (AoI). These requests have been commonly classified as one time sensing or continuous sensing requests. In the former, one-time reading from the devices of the recruited participants is needed to answer the request, while in the latter, readings are needed over a given time interval, making recruitment challenging, particularly when considering participants' mobility. Ideally, the process of recruiting participants for a given continuous sensing task should determine the best set of participants to answer the sensing requests, while satisfying two important constraints including (1) a given level of quality of information (QoI) and 2) within a given budget. This selection is also sensitive to parameters such as requirements of the sensing task with regards to the AoI coverage, and participants' mobility and distribution. To address this challenge, we propose a novel, stability-based group recruitment system for continuous sensing (Stable-GRS) that employs a genetic algorithm to select groups of participants considering their mobility patterns. The proposed system selects the most stable group of participants in the AoI that can achieve a certain level of QoI, where stability reflects the group's temporal and spatial availability. The process of recruitment is dynamic; it involves adding and removing participants throughout the sensing period to preserve the QoI requirement. Cooperative game theory, specifically the Shapley value, is used to reward selected workers based on their respective contribution. Simulations are conducted using real-life datasets and the results establish that our approach outperforms an individual-based recruitment system (IRS), which employs greedy algorithms to recruit participants for all key performance metrics, such as the QoI and costs.