what-when-how
In Depth Tutorials and Information
in the network will increase. Actually, an adversary maybe knows only a part of the
neighborhood of a target vertex. his results in better performance of the algorithm.
Although the algorithm has some advantage as we mentioned above, we should
extend the method to handle the case where not only vertices but also edges have
labels. Besides, we could induce some faked vertices in the process of anonymiza-
tion as well as considering the network structure preservation. We also can extend
1-neighborhood in Reference 1 to
d
-neighborhood to enhance the privacy preser-
vation.
k
-anonymize has some privacy problems because of the lack of diversity in
sensitive attributes. hus,
l
-diversity [7] is more desirable.
10.3 PrivacyPreservingBasedonTrustRelationships
Reference 8 proposes a distributed privacy preservation method based on the trust
relationships. his anonymization method makes use of the multi-hop routing
among cooperative nodes. he node, which represents a user in network, is uniquely
identified by a pseudonym and a node identifier.
As shown in Figure 10.7, there are three components in the system proposed in
Reference 8.
10.3.1 Matryoshkas
he matryoshkas is defined as a hierarchical structure of relationship between the
node (the owner of the matryoshkas) that is located on the core and other trusted
nodes on the concentric rings. he innermost ring consists of a set of nodes which
are trusted by the owner of the matryoshkas. he second ring consists of a set of
nodes which are trusted by the nodes in the first ring. Other rings are established
Trusted
Identification
Service
P2P (peer-to-peer)
substrate
Figure 10.7
System main entities: peer to peer substrate, matryoshkas, and
trusted identiication service [8]. (From Cutillo, L.A., Molva, R., and Strufe, T.
Wireless On-Demand Network Systems and Services, 2009.
Sixth International
Conference on WONS 2009
.February2-4,2009,pp.145-152.)
