Detecting Phantom Nodes in Wireless Sensor Networks Joengmin Hwang, Tian He, Yongdae Kim (ACM Infocom2007) Presenter : Justin
Main ideas Two factors: Prevent the phantom nodes from generating consistent ranging (distance) claims to multiple honest nodes. Detect phantom nodes by the proposed speculative method
Generating ranging claims If the locations of neighboring nodes are known, it is easy to generate a fake location. Without the location information of the neighboring nodes, it is hard for an attacker to generate a set of consistent ranging values (distances)
Generating ranging claims C B D A D’
Generating ranging claims C B D A D’ D’C and D’B decrease D’A increase
Generating ranging claims C B D A D’
Generating ranging claims C B D A D’ D’C and D’B increase D’A decrease
The detailed approach Definition: A set of nodes is consistent, if they can be projected on the unique Euclidean plane (in 3-D case, Euclidean space), keeping the measured distances among themselves.
The detailed approach Problem: Given a node set Nbr(v) that consists of a node v and its neighbors, and a distance set D that consists of the measured distance, denoted by Find the largest consistent subset of Nbr(v).
The detailed approach Two phases: Distance Measurement Phase Filtering Phase
Distance Measurement 1)Node v measures distance to each neighbor i 2)Node v announces the measured distance 3)Node i announces its measured distance to its neighbor j, and v collects 4)For each collected distance, if, it is included in the filtering phase
Filtering Using a graph G(V,E) to construct a consistent subset. The set V is used to contain the node v and its neighbors The set E is used to keep the edges between two nodes when the distance information between them maintains consistency.
Filtering 1)The local coordinate system L is determined by three nodes v, i, j with measured distance 2)Each node, calculating its location on L 3)Picking a pair of nodes, whose location on L are 4)Comparing the distance and ( which obtained in distance measurement phase ) 5)If, create edge e(i, j) in E 6)Choose the largest sizeof G(V,E)
Filtering
If, create edge e(i, j) in E Choose the largest sizeof G(V,E)
Filtering Node 6 is a phantom node
Filtering
Experiment results
Conclusions Pros Presenting a way to exclude the phantom nodes by projecting each nodes into a local coordinate The filtering operation is efficient Cons By using TDOA or TOA to measure distance, nodes need to be deployed at wide-space It’s not suitable for small area application