07 October 2011
The goal of in-network aggregation is to approximate real world phenomena (such as traffic situation, weather, or free parking spaces) using multi-hop vehicle-to-vehicle communication. Lossy data fusion techniques are employed in order to save bandwidth. Information reliability is threatened by information incompleteness due to network limitations, inaccuracy due to malfunctioning sensors and lossy aggregation, and dishonest nodes, which modify information to their own advantage. Existing security mechanisms provide some confidence in information reliability but pose constraints on flexibility and scalability of the underlying aggregation mechanism. In this talk, I will try to tie together existing approaches from different categories in a combined framework. Using small examples, I will outline the benefit of a combined approach and highlight open issues.