@inproceedings{oai:ipsj.ixsq.nii.ac.jp:00174819,
 author = {Akihiro, Yamada and Susumu, Ishihara and Akihiro, Yamada and Susumu, Ishihara},
 book = {Proceedings of The Ninth International Conference on Mobile Computing and Ubiquitous Networking},
 issue = {5},
 month = {Sep},
 note = {eal-Time Visual Car Navigation System proposed by Ishihara et al. is a system provides location-dependent up-to-date visual information requested by drivers via vehicular ad hoc network (VANET). In this system, when a driver enters the point of interest (POI), the vehicle issues a request message for the POI to other vehicles via VANET, and then vehicles that have up-to-date visual data of the POI send the data to the requesting vehicle. If multiple vehicles issue requests to the same POI, by aggregating such requests, the source vehicles of the data of the POI can select an effective route to send replies to the requesting vehicles. In addition, the source vehicles can continuously provide up-to-date information to the requesting vehicles instead of sending all new information to all vehicles in the VANET. To realize such effective delivery of location-dependent data to requesting vehicles, the data source vehicles and other vehicles that engage the data delivery should know the location of the requesting vehicles and use optimized route to the requesting vehicles. That is, vehicles should aggregate demands for location-dependent data. Ishihara et al. have proposed a probabilistic strategy for effectively aggregate such demands for location-dependent data using a soft-state-sketch-based technique. The data of the distribution of the sources of requests and the destination (POI) of the requests can be represented with a map showing the strength of requests from a position to a position, and it is called a demand map (Dmap). This paper proposes a specific scheme for realizing the strategy and evaluates the performance of the scheme through simulation considering realistic vehicular movement and DSRC communications. The proposed scheme includes two strategies for determining the timings for sending Dmap information and for selecting elements in Dmap information for effectively share the information. Simulation results show that the demand aggregation scheme achieves high concordance of the actual geographical distribution of sources of requests for POIs., eal-Time Visual Car Navigation System proposed by Ishihara et al. is a system provides location-dependent up-to-date visual information requested by drivers via vehicular ad hoc network (VANET). In this system, when a driver enters the point of interest (POI), the vehicle issues a request message for the POI to other vehicles via VANET, and then vehicles that have up-to-date visual data of the POI send the data to the requesting vehicle. If multiple vehicles issue requests to the same POI, by aggregating such requests, the source vehicles of the data of the POI can select an effective route to send replies to the requesting vehicles. In addition, the source vehicles can continuously provide up-to-date information to the requesting vehicles instead of sending all new information to all vehicles in the VANET. To realize such effective delivery of location-dependent data to requesting vehicles, the data source vehicles and other vehicles that engage the data delivery should know the location of the requesting vehicles and use optimized route to the requesting vehicles. That is, vehicles should aggregate demands for location-dependent data. Ishihara et al. have proposed a probabilistic strategy for effectively aggregate such demands for location-dependent data using a soft-state-sketch-based technique. The data of the distribution of the sources of requests and the destination (POI) of the requests can be represented with a map showing the strength of requests from a position to a position, and it is called a demand map (Dmap). This paper proposes a specific scheme for realizing the strategy and evaluates the performance of the scheme through simulation considering realistic vehicular movement and DSRC communications. The proposed scheme includes two strategies for determining the timings for sending Dmap information and for selecting elements in Dmap information for effectively share the information. Simulation results show that the demand aggregation scheme achieves high concordance of the actual geographical distribution of sources of requests for POIs.},
 pages = {1--2},
 publisher = {情報処理学会},
 title = {On the performance of a method for efficient aggregation of demands distribution for location-dependent information using soft-state sketch in VANET},
 volume = {2016},
 year = {2016}
}