A Misbehaving-Proof Game Theoretical Selection Approach for Mobile Crowd Sourcing

With the tremendous advances in ubiquitous computing, mobile crowd sourcing (MCS) has become an appealing part of the Internet of Things (IoT). In MCS systems, workers collect data with a certain quality, and get incentivized in return. However, MCS systems are vulnerable to misbehaving acts such as workers submitting multiple false or fake reports using multiple devices to affect the majority vote of the task. In addition, workers may try to maximize their profit by submitting multiple truthful data, a behavior that may prevent other potential workers to participate. The selection of such workers for a task has a negative impact on the decision making or the payoff of the task. Most of the current approaches aim to maximize the completion of tasks based on the reputation or the credibility of workers, but without consideration of tasks' payoff and the threat of misbehavior. In literature, a misbehaving act, where workers impersonate multiple identities using multiple devices to maliciously change the majority votes or selfishly increase their payment, has not been addressed during MCS recruitment. In this paper, a two-layer selection approach is proposed based on game-theory in which the payoff of the tasks is maximized based on the individual contributions of the workers. In addition, the proposed model detects and eliminates the misbehaving act where workers submit multiple reports using multiple devices, during the recruitment phase. Simulations using real-life datasets show that the proposed approach succeeds in detecting and eliminating misbehaving devices, and outperforms the benchmarks in terms of the payoff of the tasks.