Trust Management of Encrypted Data Aggregation in a Sensor Network Environment

Abstract

Efficient power management is vital for increasing the life of sensor nodes in wireless sensor networks (WSN). Thus, techniques such as data aggregation have been widely used in WSN to preserve nodes' energy. Despite its appealing and powerful features, data aggregation processes require a high level of security. Ensuring accuracy with aggregating data can become an issue even due to single bit errors. Hence, many encryption techniques have been proposed to secure the data aggregation process. However, once a wireless sensor is physically compromised, the keys stored on the sensor are exposed, thus rendering the cryptography useless. Moreover, most cryptographic approaches restrict the level of collaboration with other security measures -- especially trust management, where behavior observation is vital. In this paper, we propose a trust management approach on encrypted packets to deal with the faulty nodes in sensor networks. We designed a collaborative framework which incorporates the Order-Preserved Encryption Scheme [2] and Sigmoid Trust Management approaches. We then discuss how these two can coexist and comprehensively evaluate an integrated OPES-trust management scheme in the sensing environment using both the TOSSim simulator and an actual mote implementation. The results have shown that trust management and OPES cryptography does not consume more energy than radio transmission, yet provides a fairly accurate result.