Towards efficient detection of sybil attacks in location-based social networks
A location-based social network (LBSN) can facilitate interactions between nearby strangers. However, it is vulnerable to a Sybil attack, in which an attacker can subvert the reputation system of the LBSN by manipulating trust relationships among users. Such an attack cannot be detected using the existing community analysis-based countermeasures designed for regular social networks, since both the Sybil communities and the non-Sybil communities are quite sparse and rarely connected in LBSNs. By carefully analyzing honest user trajectories and Sybil attack behavior, we observe that in LBSNs, a Sybil attack usually comes together with continuous user gatherings. Based on this observation, we design a Bloom filter-based user gathering detection scheme, which can effectively and efficiently detect Sybil attacks in LBSNs. Experimental evaluation using real-world user location data validates the effectiveness and efficiency of our scheme.
2017 IEEE Symposium Series on Computational Intelligence (SSCI)
Towards efficient detection of sybil attacks in location-based social networks.
2017 IEEE Symposium Series on Computational Intelligence (SSCI), 1-7.
Retrieved from: https://digitalcommons.mtu.edu/cs_fp/8