Clustering and sensing with decentralized detection in vehicular ad hoc networks

摘要: Abstract In the near future, vehicles will be more and advanced sensing platforms: for instance, at least one smartphone (with several on-board sensors) is likely to inside each vehicle. Smartphone-based inter-vehicle communications thus support creation of vehicular sensor networks (VSNs). this paper, we analyze performance clustered VSNs, where (hierarchical) decentralized detection schemes are used estimate status an observed spatially constant phenomenon interest. Clustering makes processing efficient architecture scalable. Our approach consists creation, during a downlink phase, VSN topology through fast broadcast control messages, started from remote sink (e.g., in cloud), novel clustering protocol, denoted as cluster-head election irresponsible forwarding (CEIF). This then exploited, uplink collect sensed data perform distributed detection. The proposed scheme investigated considering mostly IEEE 802.11b (smartphone-based) well 802.11p (inter-vehicle) both highway-like urban-like scenarios. results highlight existing trade-off between decision delay energy efficiency. VSN-based have cope with “ephemeral” nature clusters. Therefore, proper cluster maintenance strategies needed prolong lifetime and, consequence, maximum amount which can collected before clusters break. leads concept “on move.”