Distributed Privacy-preserving Solutions to k-Anonymity Problem in Wireless Sensor Networks

As important as other security issues, the contextual privacy problem is notable but not thoroughly addressed in distributed networking systems, such as Wireless Sensor Networks(WSNs). An adversary can launch contextual privacy attacks via assigning the linking probability of users’ identities and target messages during data transmissions, to further acquire the sensitive relationship information in data-centric networks. By segmenting the whole anonymity set into subgroups, a superior contextual privacy attack called Multiplex Uniform Attack (MUA) is identified as the first work of this kind and investigated in this paper. To reach the privacy expectation of k-anonymity in WSNs, we propose a novel mask-free privacy protection mechanism called k-Anonymous Transfer Mix (ATM) against MUA. In addition, three token-based distributed algorithms are designed to defend against this privacy attack in fixed, adaptive and lightweight ways according to diversified data metrics. Theoretical analysis demonstrates the effectiveness of our proposed mechanism on privacy protection and experimental simulations validate the theoretical results.