“End-by-Hop” Data Integrity

Abstract

Wireless sensor networks have been proposed for various applications, such as environmental monitoring and tactical military applications. For most of these applications sensors, scattered across a large physical area, organize themselves into a network that forwards data back to a central sink. Sensors are typically assumed to be severely constrained with respect to energy consumption, computational power and communication capabilities (especially the data rate and range of the transmitter). Data-centric networking, where forwarding nodes aggregate or filter data en-route to the central sink, have been proposed to reduce the amount of data transported in the network and conserve energy. This means that data-centric networks are significantly different from traditional end-to-end networks, because data are altered on every hop from the source to the sink.

Traditional end-to-end integrity mechanisms ensure that data cannot be modified on the way from the source to the destination. In data-centric networking, however, data is supposed to be altered on every hop from the source to the sink, so new integrity mechanisms must be investigated. In this paper we propose a new “end-by-hop” data integrity service that supports aggregation or filtering in data-centric networks. We also describe a mechanism that could be used to provide this service and provide an initial analysis of the efficiency and security of the mechanism proposed. One of the desirable properties of the proposed mechanism is that it allows the system architect to trade-off the computational load on the sensor nodes against a higher computational load at the sink, which we assume does not have the same severe resource limitations as the sensor nodes.