A lightweight privacy-preserving truth discovery in mobile crowdsensing systems

Truth discovery as an effective method to improve data quality in mobile crowd sensing has recently gained widespread attention. It inferred participant weight based on the sensory data submitted by participants, and then used the weight to aggregate sensory data and finally inferred the real information. Due to participants in mobile crowd sensing facing the problem of privacy leakage, existing work mainly focuses on sensory data privacy, with less consideration of weight privacy. Based on this, this paper proposes a lightweight privacy-preserving truth discovery in mobile crowd sensing ALPPTD. ALPPTD ran the encryption and decryption calculations of weight and truth update on the cloud server side, which greatly reduced the computation overhead of participants to motivate more users to participate. Meanwhile, two non-colluding cloud servers use homomorphic encryption to achieve aggregation of sensory data, thus iteratively computing the truth while guaranteeing the privacy of participants’ sensory data and weights. Theoretical analysis and experiment results show that ALPPTD ensures the privacy of participants’ sensory data and weight while computing the truth value with low computation overhead characteristics of participants.