Secure Collaborative Spectrum Sensing In Cognitive Radio Networks

Cognitive Radio (CR) has been regarded as the promising approach to address the spectrum shortage problems. It requires the secondary users to detect the activity of the primary user, and allows them to access the idle spectrum. In CR networks, secondary users are often supposed to collaboratively sense the spectrum, since this cooperation can significantly improve the spectrum sensing accuracy, due to the spatial diversityHowever, collaborative spectrum sensing may also incur security issues.In this paper, we identify two new security threats in collaborative sensing: Identify the Location Privacy Leakage in collaborative spectrum sensing. We discover that when the secondary user participates in the collaboration, his location information is also leaked. Malicious external attacker or un-trusted Fusion Center (FC) can achieve involuntary geo-location of a secondary user by linking his location-dependent sensing report to his physical position.Identify Entropy Selfishness in collaborative spectrum sensing. Different from traditional selfishness models which mainly focus on the free-riding during report sharing procedure, ES arises from spectrum sensing process. In particular, the ES user doesn't make his own sensing, and just claims a duplicated or slightly modified report from others as his own.Further, we propose the Privacy Preserving collaborative Spectrum Sensing (PPSS) scheme to protect the secondary user's location privacy. This scheme includes two part: 1) Privacy Preserving Sensing Report Aggregation (PPSRA) protocol, and 2) Distributed Dummy Report Injection (DDRI) protocol. PPSRA can conceal the secondary user's sensing report in aggregation, while DDRI can further protect the user's differential location privacy. Via experiment, we demonstrate the effectiveness and practicality of your scheme.At the same time, we propose Puzzle based Message Masking (PMM) scheme to stimulate the entropy selfish user to sense the spectrum. We design a non-cryptographic puzzle construction, which can prevent entropy selfish user from utilizing the shared sensing reports. Meanwhile, for the honest user, sensing the spectrum enable him to solve this puzzle and therefore utilize the shared sensing reports. Thus, this puzzle can stimulate the entropy selfish user to sense the spectrum. Finally, we demonstrate the effectiveness of PMM by experiment.