Replica voting based architectures for reliable data dissemination in vehicular networks

Vehicular networks often consist of computational devices that collect data representing the external environment (e.g., road traffic volume, terrain conditions, and vehicle tracking). The data collected is prone to errors for two reasons: first, the processing algorithms in devices often have only limited capabilities that result in a fuzzy and imprecise representation of the external datum; second, some of the devices may be maliciously faulty mis-reporting the external datum. In such a setting, replicating the devices and voting on the environment data collected by them enhances the trust-worthiness of data reported to the application. The paper describes a cooperation mechanism that embodies two complementary functions: replication of devices and voting among devices. Device-level cooperation is realized using a tree-structured overlay set up over a vehicular network, in which the root node is attached to a data dissemination station and the leaf nodes are attached to the data collection devices in different geographic regions. Voting among the data collected is carried out at the leaf nodes whereupon a sanitized data propagates up the tree towards the root. An intermediate node may also carry out aggregation functions on the data arriving from its downstream tree segments and then forward the aggregated data upstream. The paper describes the voting-based data collection and dissemination as a middleware service.