Burgstahler, Daniel-Maurice (2017)
Collaborative Sensing in Automotive Scenarios : Enhancement of the Vehicular Electronic Horizon through Collaboratively Sensed Knowledge.
Technische Universität Darmstadt
Dissertation, Erstveröffentlichung
Kurzbeschreibung (Abstract)
Modern vehicles are equipped with a variety of advanced driver assistance systems that increase driving comfort, economy and safety. Respective information sources for these systems are local sensors, like cameras, radar or lidar. However, the next generation of assistant systems will require information above the local sensing range. An extension of the local perception can be provided by the use of appro- priate communication mechanisms. Hence, other vehicles can serve as an informa- tion source by providing their local perception data, but also any other information source, such as cloud services. Required communication can take place directly be- tween vehicles via mobile ad-hoc communication or via a backend by the use of cellu- lar communication. The appropriate technology depends on the respective use case, that determines information content, granularity and tolerated latency. Based on liter- ature, we derived a categorization of use case dependent information demands, with respect to communication. The resulting three zones, namely safety zone, awareness zone and information zone, refer to the tolerated latency between the occurrence of an information and the point in time the information has to be processed at the receiver side. While communication mechanisms for the safety zone, i. e., the ego-vehicle’s di- rect surroundings with a remaining driving time of less than 2 − 5 seconds, have been focus in research and standardization in the past, respective mechanisms for larger distances have not been sufficiently considered. In this thesis, we examine in- formation distribution mechanisms in context of the previously mentioned use case categories. As the first key contribution, we consider the gathering of vehicular sensed data with regard to the information zone, i. e., more than 30 seconds remaining driving time to the point of the information origin. We developed a probabilistic data collection model that is able to reduce data traffic up to 85 % compared to opportunistic trans- mission and still sticks to certain quality metrics, e. g., a maximum detection latency. A central adaption of transmission probabilities to the density of transmitting vehi- cles is applicable for cellular use and copes with sparse traffic situations. Moreover, we have extended this approach by hybrid communication, i. e., the parallel use of cellular and mobile ad-hoc communication. This allows to further reduce cellular based data traffic, in particular in case of dense traffic. As the second key contribution, we examine the efficient distribution of the pre- viously gathered information. Information is structured and prioritized according to the most probable driving path, as so-called electronic horizon. The transmission towards the vehicles is performed in small data packets, according to the given pri- orities. The aim is to transmit only information relevant for road segments that will be used. Concerning this, we developed a mechanism for most probable travel path estimation and a data structure for efficient mapping of the electronic horizon. As the third key contribution, we examine the information exchange in the aware- ness zone, an area between the safety zone and the information zone with about 5 to 30 seconds remaining driving time to the point of the information origin. Derived from the respective use cases, this data is not directly safety relevant, but it is still about dynamic position information of neighboring vehicles. Due to the relatively
long distance, direct vehicle to vehicle communication is not possible. Respective data has to be forwarded by intermediate vehicles. However, position beacons with- out data forwarding can already cause channel congestion in dense traffic situations. The use of cellular networks would require absolute total network coverage with permanent free channel resources. To enable forwarding of dynamic vehicle infor- mation anyhow, we developed at first a mechanism to reduce the channel load for position beacons. Next, we use the freed-up bandwidth to forward dynamic informa- tion about neighboring vehicle positions. With this mechanism, we are able to more than double the range of vehicular perception, with respect to moving objects. In extension to standardized communication mechanisms for the safety relevant direct proximity, our three mentioned contributions provide the means to complete the long range vehicular perception for future advanced driver assistance systems.
Typ des Eintrags: | Dissertation | ||||
---|---|---|---|---|---|
Erschienen: | 2017 | ||||
Autor(en): | Burgstahler, Daniel-Maurice | ||||
Art des Eintrags: | Erstveröffentlichung | ||||
Titel: | Collaborative Sensing in Automotive Scenarios : Enhancement of the Vehicular Electronic Horizon through Collaboratively Sensed Knowledge | ||||
Sprache: | Englisch | ||||
Referenten: | Steinmetz, Prof. Dr. Ralf ; David, Prof. Dr. Klaus | ||||
Publikationsjahr: | 2017 | ||||
Ort: | Darmstadt | ||||
Datum der mündlichen Prüfung: | 19 Juni 2017 | ||||
URL / URN: | http://tuprints.ulb.tu-darmstadt.de/6554 | ||||
Kurzbeschreibung (Abstract): | Modern vehicles are equipped with a variety of advanced driver assistance systems that increase driving comfort, economy and safety. Respective information sources for these systems are local sensors, like cameras, radar or lidar. However, the next generation of assistant systems will require information above the local sensing range. An extension of the local perception can be provided by the use of appro- priate communication mechanisms. Hence, other vehicles can serve as an informa- tion source by providing their local perception data, but also any other information source, such as cloud services. Required communication can take place directly be- tween vehicles via mobile ad-hoc communication or via a backend by the use of cellu- lar communication. The appropriate technology depends on the respective use case, that determines information content, granularity and tolerated latency. Based on liter- ature, we derived a categorization of use case dependent information demands, with respect to communication. The resulting three zones, namely safety zone, awareness zone and information zone, refer to the tolerated latency between the occurrence of an information and the point in time the information has to be processed at the receiver side. While communication mechanisms for the safety zone, i. e., the ego-vehicle’s di- rect surroundings with a remaining driving time of less than 2 − 5 seconds, have been focus in research and standardization in the past, respective mechanisms for larger distances have not been sufficiently considered. In this thesis, we examine in- formation distribution mechanisms in context of the previously mentioned use case categories. As the first key contribution, we consider the gathering of vehicular sensed data with regard to the information zone, i. e., more than 30 seconds remaining driving time to the point of the information origin. We developed a probabilistic data collection model that is able to reduce data traffic up to 85 % compared to opportunistic trans- mission and still sticks to certain quality metrics, e. g., a maximum detection latency. A central adaption of transmission probabilities to the density of transmitting vehi- cles is applicable for cellular use and copes with sparse traffic situations. Moreover, we have extended this approach by hybrid communication, i. e., the parallel use of cellular and mobile ad-hoc communication. This allows to further reduce cellular based data traffic, in particular in case of dense traffic. As the second key contribution, we examine the efficient distribution of the pre- viously gathered information. Information is structured and prioritized according to the most probable driving path, as so-called electronic horizon. The transmission towards the vehicles is performed in small data packets, according to the given pri- orities. The aim is to transmit only information relevant for road segments that will be used. Concerning this, we developed a mechanism for most probable travel path estimation and a data structure for efficient mapping of the electronic horizon. As the third key contribution, we examine the information exchange in the aware- ness zone, an area between the safety zone and the information zone with about 5 to 30 seconds remaining driving time to the point of the information origin. Derived from the respective use cases, this data is not directly safety relevant, but it is still about dynamic position information of neighboring vehicles. Due to the relatively long distance, direct vehicle to vehicle communication is not possible. Respective data has to be forwarded by intermediate vehicles. However, position beacons with- out data forwarding can already cause channel congestion in dense traffic situations. The use of cellular networks would require absolute total network coverage with permanent free channel resources. To enable forwarding of dynamic vehicle infor- mation anyhow, we developed at first a mechanism to reduce the channel load for position beacons. Next, we use the freed-up bandwidth to forward dynamic informa- tion about neighboring vehicle positions. With this mechanism, we are able to more than double the range of vehicular perception, with respect to moving objects. In extension to standardized communication mechanisms for the safety relevant direct proximity, our three mentioned contributions provide the means to complete the long range vehicular perception for future advanced driver assistance systems. |
||||
Alternatives oder übersetztes Abstract: |
|
||||
URN: | urn:nbn:de:tuda-tuprints-65541 | ||||
Sachgruppe der Dewey Dezimalklassifikatin (DDC): | 000 Allgemeines, Informatik, Informationswissenschaft > 004 Informatik 600 Technik, Medizin, angewandte Wissenschaften > 620 Ingenieurwissenschaften und Maschinenbau |
||||
Fachbereich(e)/-gebiet(e): | 18 Fachbereich Elektrotechnik und Informationstechnik 18 Fachbereich Elektrotechnik und Informationstechnik > Institut für Datentechnik > Multimedia Kommunikation 18 Fachbereich Elektrotechnik und Informationstechnik > Institut für Datentechnik |
||||
Hinterlegungsdatum: | 23 Jul 2017 19:56 | ||||
Letzte Änderung: | 23 Jul 2017 19:56 | ||||
PPN: | |||||
Referenten: | Steinmetz, Prof. Dr. Ralf ; David, Prof. Dr. Klaus | ||||
Datum der mündlichen Prüfung / Verteidigung / mdl. Prüfung: | 19 Juni 2017 | ||||
Export: | |||||
Suche nach Titel in: | TUfind oder in Google |
Frage zum Eintrag |
Optionen (nur für Redakteure)
Redaktionelle Details anzeigen |