Protecting Storage Location Privacy in Sensor Networks

Numerous schemes have been proposed to facilitate data collection and provision in sensor networks, among which the Data-Centric Storage (DCS) scheme is an energy-efficient solution and a popular choice for many sensor network applications. However, since each sensor node in the DCS system knows the locations of all storage nodes, the DCS system is extremely vulnerable to security attacks as a single compromised sensor node will expose all the storage locations to the adversary. To address this problem, we propose a randomized storage concealment scheme along with a supplementary storage migration scheme. In the randomized storage concealment scheme, sensor nodes cooperate to forward data towards the storage nodes without keeping explicit storage locations; instead, each sensor node only maintains the IDs of its randomly-picked next-hop nodes towards the storage nodes. This scheme increases the difficulty significantly for the adversary to derive the storage locations. Nevertheless, the protection provided by this scheme degrades gradually as more and more sensor nodes are compromised. Hence, we further introduce a storage migration scheme to supplement the randomized storage concealment scheme, which directs the storage duties to migrate periodically among sensor nodes. Extensive analysis and simulations are conducted to show that the proposed schemes can effectively protect the storage location privacy with modestly added overhead.