In-network aggregation is an essential data-processing primitive in sensor networks. Sensor nodes forward data towards the sink (i.e. the base station which will forward the collected data to the querier). As sensor nodes closer to the sink receive data from nodes further away, they aggregate the information into concise digests. This results in significant energy savings over having each node forward their respective readings directly to the sink.

However, if malicious nodes are present in the network, then their influence on the collected data is siggnificantly magnified by the use of in-network data aggregation. For example, an aggregator node could claim that all the nodes downstream from it reported a very large value. In this way, the compromise of just a few aggregator nodes could have an unbounded effect on the correctness of the data returned by a sensor network.

In this project we develop algorithms that can improve the integrity guarantees of the result returned by a sensor network where malicious nodes may be present. The algorithms we develop allow the querier to detect if a returned result was falsified by a malicious aggregator and thus reject it.

Contact Us