The Feasibility of a Search Engine for Metropolitan Vehicular Ad-Hoc Networks

Applications based on vehicular ad-hoc networks (VANETs) rely on information exchanged within the network. There are two different methods of exchanging information among vehicles, i.e. distributing it proactively or requesting it on-demand. Proactive distribution of information is suitable for distributing small amounts of data, comprising information of common interest for a large number of vehicles. In contrast, proactive distribution of data is not advisable if information contained therein is relevant only to a small number of vehicles or if it would require large amounts of data. A more efficient alternative in this case is on-demand information distribution, which allows information of interest to be located, requested and sent on-demand. On-demand information retrieval in sparse metropolitan-wide VANETs is challenging, as vehicles stay in the metropolitan area only for a short period of time and message exchange is impacted by communication delays. In this paper, we introduce three different concepts of a search engine for metropolitan VANETs and explain how these concepts could make it possible to allow on-demand information exchange. We compare the concepts for their applicability by evaluating their performance with respect to latency and network load.