Tobias Meuser

Adaptive Decision Making Based on Temporal Information Dynamics

To increase road safety and efficiency, connected vehicles rely on the exchange of information. On each vehicle, a decision-making algorithm processes the received information and determines the actions that are to be taken. State-of-the-art decision approaches focus on static information and ignore the temporal dynamics of the environment, which is characterized by high change rates in a vehicular scenario. Hence, they keep outdated information longer than necessary and miss optimization potential. To address this problem, we propose a quality of information (QoI) weight based on a Hidden Markov Model for each information type, e.g., a road congestion state. Using this weight in the decision process allows us to combine detection accuracy of the sensor and the information lifetime in the decision-making, and, consequently, adapt to environmental changes significantly faster. We evaluate our approach for the scenario of traffic jam detection and avoidance, showing that it reduces the costs of false decisions by up to 25% compared to existing approaches.

ProbSense.KOM: A Probabilistic Sensing Approach for Gathering Vehicular Sensed Data

Advanced driver assistance systems (ADAS) improve safety, energy efficiency and driver comfort. Such systems are commonly based on sensor data, however, sensor range is physically limited. A way to extend the sensing range is to share sensor reading with others, i.e., vehicles and infrastructure services. Since direct vehicle communication is not widely deployed and vehicles are often not driving in direct communication range, communication has to be realized via cellular networks. Due to high costs for cellular communication, the transmission of sensed data has to be efficient and the amount of transmitted data must be minimized. As possible solution, we introduce a concept of probabilistic data transmission for vehicular sensed data. The system divides the map into geographic cells, and a probabilistic model is managed for each geographic cell individually. We are able to achieve a reduction in data transmission volume of up to 50% in comparison to opportunistic approaches.

Conducting a Large-scale Field Test of a Smartphone-based Communication Network for Emergency Response

Smartphone-based communication networks form a basis for services in emergency response scenarios, where communication infrastructure is impaired or overloaded. Still, their design and evaluation are largely based on simulations that rely on generic mobility models and weak assumptions regarding user behavior. For a realistic assessment, scenario-specific models are essential. To this end, we conducted a large-scale field test of a set of emergency services that relied solely on ad hoc communication. Over the course of one day, we gathered data from smartphones distributed to 125 participants in a scripted disaster event. In this paper, we present the scenario, measurement methodology, and a first analysis of the data. Our work provides the first trace combining user interaction, mobility, and additional sensor readings of a large-scale emergency response scenario, facilitating future research.

Cellular Bandwidth Prediction for Highly Automated Driving - Evaluation of Machine Learning Approaches based on Real-World Data.

To enable highly automated driving and the associated comfort services for the driver, vehicles require a reliable and constant cellular data connection. However, due to their mobility vehicles experience significant fluctuations in their connection quality in terms of bandwidth and availability. To maintain constantly high quality of service, these fluctuations need to be anticipated and predicted before they occur. To this end, different techniques such as connectivity maps and online throughput estimations exist. In this paper, we investigate the possibilities of a large-scale future deployment of such techniques by relying solely on lowcost hardware for network measurements. Therefore, we conducted a measurement campaign over three weeks in which more than 74,000 throughput estimates with correlated network quality parameters were obtained. Based on this data set—which we make publicly available to the community—we provide insights in the challenging task of network quality prediction for vehicular scenarios. More specifically, we analyse the potential of machine learning approaches for bandwidth prediction and assess their underlying assumptions.

Ad-Hoc-Kommunikation – Gesellschaftlich wünschenswert, rechtlich ungeregelt

Bei einem Ad-hoc-Kommunikationsnetzwerk auf Peer-to-Peer-Basis ist rechtlich gesehen jede Nutzerin und jeder Nutzer zugleich Anbieter. Dieser hat aber keinerlei Moglichkeit, auf die Technik selbst Einfluss zu nehmen und verfugt auch uber keine vollstandigen Informationen uber das Netzwerk. Sogar die Gesamtzahl der Teilnehmerinnen und Teilnehmer ist der einzelnen Nutzerin oder dem einzelnen Nutzer in der Regel unbekannt. Dadurch ist es ihm unmoglich, seine datenschutzrechtlichen Pflichten zu erfullen. Als Losungsansatz bietet sich hier die Verpflichtung des App-Anbieters an. Hierzu bedarf es spezieller gesetzlicher Regelungen fur diese Kommunikationsnetzwerke, damit diese ihren gesellschaftlich wunschenswerten Effekt auch verwirklichen konnen.