Trust-Aware sensing Quality estimation for team Crowdsourcing in social IoT

In the Internet of Things (IoT), the mobile smart devices with powerful sensing capability help mobile crowdsourcing become an important paradigm to sense environment information. The social Internet of Things paradigm can be exploited for complex task crowdsourcing by forming a collaborative team of socially connected nodes (i.e., smart devices). Few existing team crowdsourcing studies have ever satisfied requirements of trustworthy sensing data and collaborative communication among team members. In this paper, we design TAQ-Crowd (Trust-Aware sensing Quality estimation for team Crowdsourcing), a social team crowdsourcing framework for Social Internet of Things systems. Within TAQ-Crowd, we first incorporate the consideration of trustworthy relationships between nodes into sensing data quality evaluation for TAQ model design. Then, we design a task assignment algorithm CS-Selection, in which the sensing quality guides the participant selection to maximize the overall task valuation under a budget constraint. Meanwhile, we consider a variant of the classic Traveling Salesman Problem (TSP) to extract a tree-structured routing network for team communication. Solving the team crowdsourcing problem concerns participating device selection and task cooperation, which involves two coupling NP-hard problems. The two coupling problems can be transformed into an essentially submodular cost submodular knapsack problem to be solved by the greedy task assignment strategy. Finally, extensive simulation experiments are conducted. The results show that TAQ-Crowd significantly outperforms state-of-art approaches in team formation with at least (1 - 1/e)/2 approximation ratio. Furthermore, the achieved superior performance can validate our proposed trust-aware sensing quality estimation.