Privacy-Preserving Collaborative Spectrum Sensing With Multiple Service Providers

Abstract

In cognitive radio networks (CRNs), collaborative sensing has been considered an attractive means to improving spectrum sensing performance. However, privacy issues arise when multiple service providers (SPs) collaborate on learning the spectrum availabilities. Specifically, sharing sensing data may enable malicious SPs or secondary users (SUs) to geolocate an SU using existing localization techniques. These malicious entities could be untrusted SPs/SUs or external attackers that compromise SPs/SUs. To incentivize SUs to contribute their sensing data, the privacy of each SU should be guaranteed. In this paper, we propose a privacy preservation framework called PrimCos for multi-SP collaborative sensing, which addresses several competing challenges not yet considered in the literature, that is, being compatible with general collaborative sensing schemes, providing privacy guarantee for each SU and ensuring worst case privacy protection under collusion. Both analytical and numerical results show that the proposed framework provides privacy protection for each SU with controllable impact on the sensing performance under different types of attacks.