Dynamic Energy-based Encoding and Filtering in Sensor Networks

In critical sensor deployments it is important to ensure the authenticity and integrity of sensed data. Further, one must ensure that false data injected into the network by malicious nodes is not perceived as accurate data. In this paper we present the Dynamic Energy-based Encoding and Filtering (DEEF) framework to detect the injection of false data into a sensor network. DEEF requires that each sensed event report be encoded using a simple encoding scheme based on a keyed hash. The key to the hashing function dynamically changes as a function of the transient energy of the sensor, thus requiring no need for re-keying. Depending on the cost of transmission vs. computational cost of encoding, it may be important to remove data as quickly as possible. Accordingly, DEEF can provide authentication at the edge of the network or authentication inside of the sensor network. Depending on the optimal configuration, as the report is forwarded, each node along the way verifies the correctness of the encoding probabilistically and drops those that are invalid. We have evaluated DEEF's feasibility and performance through analysis. Our results show that DEEF, without incurring transmission overhead (increasing packet size), is able to eliminate 90% - 99% of false data injected from an outsider within 9 hops before it reaches the sink.