School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON K1N 6N5, Canada.
Watts Water Technologies, North Andover, MA 01845, USA.
Sensors (Basel). 2020 Apr 6;20(7):2055. doi: 10.3390/s20072055.
Mobile CrowdSensing (MCS) is an emerging paradigm in the distributed acquisition of smart city and Internet of Things (IoT) data. MCS requires large number of users to enable access to the built-in sensors in their mobile devices and share sensed data to ensure high value and high veracity of big sensed data. Improving user participation in MCS campaigns requires to boost users effectively, which is a key concern for the success of MCS platforms. As MCS builds on non-dedicated sensors, data trustworthiness cannot be guaranteed as every user attains an individual strategy to benefit from participation. At the same time, MCS platforms endeavor to acquire highly dependable crowd-sensed data at lower cost. This phenomenon introduces a game between users that form the participant pool, as well as between the participant pool and the MCS platform. Research on various game theoretic approaches aims to provide a stable solution to this problem. This article presents a comprehensive review of different game theoretic solutions that address the following issues in MCS such as sensing cost, quality of data, optimal price determination between data requesters and providers, and incentives. We propose a taxonomy of game theory-based solutions for MCS platforms in which problems are mainly formulated based on Stackelberg, Bayesian and Evolutionary games. We present the methods used by each game to reach an equilibrium where the solution for the problem ensures that every participant of the game is satisfied with their utility with no requirement of change in their strategies. The initial criterion to categorize the game theoretic solutions for MCS is based on co-operation and information available among participants whereas a participant could be either a requester or provider. Following a thorough qualitative comparison of the surveyed approaches, we provide insights concerning open areas and possible directions in this active field of research.
移动群感（MCS）是智能城市和物联网（IoT）数据分布式获取中的一种新兴范例。MCS 需要大量用户来启用对其移动设备中内置传感器的访问，并共享感测数据，以确保大数据的高价值和高精度。提高用户对 MCS 活动的参与度需要有效地激励用户，这是 MCS 平台成功的关键关注点。由于 MCS 基于非专用传感器，因此不能保证数据的可信度，因为每个用户都采用了从参与中受益的个人策略。同时，MCS 平台努力以较低的成本获取高度可靠的众感数据。这种现象在用户之间以及参与者池和 MCS 平台之间形成了博弈。针对各种博弈论方法的研究旨在为该问题提供稳定的解决方案。本文全面回顾了不同的博弈论解决方案，这些解决方案解决了 MCS 中的以下问题，例如感测成本、数据质量、数据请求者和提供者之间的最优价格确定以及激励措施。我们提出了一种基于博弈论的 MCS 平台解决方案分类法，其中问题主要基于 Stackelberg、贝叶斯和进化博弈来制定。我们介绍了每种博弈达到均衡的方法，其中问题的解决方案确保了博弈的每个参与者都对其效用感到满意，而无需改变其策略。对 MCS 的博弈论解决方案进行分类的初始标准是基于参与者之间的合作和可用信息，而参与者可以是请求者或提供者。对调查方法进行了彻底的定性比较之后，我们就该活跃研究领域的开放领域和可能的方向提供了一些见解。
