Bilateral Personalized Information Fusion in Mobile Crowdsensing

In mobile crowdsensing, information fusion is an important process to achieve data aggregation results. In recent years, people have paid more and more attention to satisfying the personalized requirements of users in information fusion. Bilateral users at both ends of the crowdsensing server have different personalized requirements due to their different roles in crowdsensing. However, existing personalized mechanisms often only focus on satisfying the requirements of data contributors, often ignoring the potential personalized requirements of data consumers. Furthermore, to the best of our knowledge, there is no previous research on how to provide personalized data aggregation results for data consumers with different data quality requirements. To address these challenges, this paper proposes a novel bilateral personalized information fusion mechanism, called BPIF, to provide personalized data aggregation results for data consumers with different data quality requirements while ensuring privacy protection for data contributors. In our approach, we design a bilateral personalized information fusion scheme based on the personalized sampling mechanism in mobile crowdsensing. We obtain the final data set through double rounds of sampling. After aggregating the final data set, we use the differential privacy mechanism to perturb the aggregation results to protect the privacy information of data contributors. To verify the feasibility and effectiveness of BPIF, we conduct experiments on virtual and real-world datasets, demonstrating its potential advantages in mobile crowdsensing environments.