13 results
Now showing1 - 10 of 13
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Antenna setup for future joint radar-communications - characteristics and mounting positions(Copernicus Publications, 2023-03-21); ; ; ; The development of millimeter wave systems is driven by the strong trend toward new communications generations and especially by the emerging joint radar and communications design approach. Safety-critical applications like platooning or intersection assistance will significantly benefit from the combination of sensing and communications. While radar performs a channel measurement and thus, needs a wide field of view (especially in city/intersection scenarios), communications aims to minimize the interference for other not addressed receivers (e. g. in a platoon) by a focused antenna design. The proposed work extends the analysis of the influence of various antenna positioning for a typical automotive scene by taking also different characteristics (antenna gain, half power beamwidth, and sidelobe level) into account. Hereby, it is mandatory to investigate the communications and sensing performance simultaneously. The positions at the front bumper - typical for radar sensors - and especially at the left mirror convinced regarding the vehicular communications as well as the sensing behaviour. Applying focused antennas is promising, however, has also limits if the signals are not received out of the main beam but out of the sidelobes, resulting in a critical communications performance. Thus, beam steering is recommended to be applied in the future.Publicationtype: Journal ArticleScopus© Citations 1 75 94 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , HARadNet: Anchor-free target detection for radar point clouds using hierarchical attention and multi-task learningTarget localization and classification from radar point clouds is a challenging task due to the inherently sparse nature of the data with highly non-uniform target distribution. This work presents HARadNet, a novel attention based anchor free target detection and classification network architecture in a multi-task learning framework for radar point clouds data. A direction field vector is used as motion modality to achieve attention inside the network. The attention operates at different hierarchy of the feature abstraction layer with each point sampled according to a conditional direction field vector, allowing the network to exploit and learn a joint feature representation and correlation to its neighborhood. This leads to a significant improvement in the performance of the classification. Additionally, a parameter-free target localization is proposed using Bayesian sampling conditioned on a pre-trained direction field vector. The extensive evaluation on a public radar dataset shows an substantial increase in localization and classification performance.Publicationtype: Journal Article47 10 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , A machine learning perspective on automotive radar direction of arrival estimation(IEEE, 2022-01-07); ; ; ; Millimeter-wave sensing using automotive radar imposes high requirements on the applied signal processing in order to obtain the necessary resolution for current imaging radar. High-resolution direction of arrival estimation is needed to achieve the desired spatial resolution, limited by the total antenna array aperture. This work gives an overview of the recent progress and work in the field of deep learning based direction of arrival estimation in the automotive radar context, i.e. using only a single measurement snapshot. Additionally, several deep learning models are compared and investigated with respect to their suitability for automotive angle estimation. The models are trained with model- and data-based approaches for data generation, including simulated scenarios as well as real measurement data from more than 400 automotive radar sensors. Finally, their performance is compared to several baseline angle estimation algorithms like the maximum-likelihood estimator. All results are discussed with respect to the estimation error, the resolution of closely spaced targets and the total estimation accuracy. The overall results demonstrate the viability and advantages of the proposed data generation methods, as well as super-resolution capabilities of several architectures.Publicationtype: Journal ArticleScopus© Citations 29 117 1620 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , High-Resolution Direction-of-Arrival Estimation using Distributed Radar SensorsWe present an approach to high-resolution direction-of-arrival estimation for distributed radar sensors without the need for synchronization or wired connections. The sensor signals are post-processed and combined after range-Doppler processing achieving a larger virtual aperture and multiple snapshots. The proposed approach is validated via simulations and the effectiveness is compared to other baseline estimators with respect to estimation accuracy and resolution performance. Finally, the applicability on real data is demonstrated using a simple measurement setup in an anechoic chamber.Publicationtype: Conference Paper48 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Bayesradar : Bayesian Metric-Kalman filter learning for improved and reliable radar target classification(IEEE, 2021-10-25); ; ; ; Publicationtype: Conference Paper83 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Evaluation of Different Antenna Positions for Joint Radar-Communication at 77 GHz(2021-09-28); ; ; ; The development of millimeter wave systems is driven by the strong trend towards the new communication generations and especially by the emerging joint radar and communication design approach. Safety critical applications like platooning or intersection assistance will significantly benefit from the combination of sensing and communication. This combination requires that both operate on the same frequency bands either in a co-design, a cooperation or even in a coexistence, using the same hardware. However, each technology has its own requirements. While radar performs a channel measurement and thus, needs a wide field of view, communication aims to minimize the interference for other not addressed receivers (e. g. in a platoon) by a focused antenna design. The proposed work analyses the effect of antenna positioning and antenna characteristics (omnidirectional and focused) for a typical automotive scene by evaluating both the communication and sensing performance. Especially, the positions at the front bumper - typical for radar sensors- and at the left mirror convinced regarding the vehicular communication as well as the sensing behaviour.Publicationtype: Conference Paper68 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Validation and Analysis of the Propagation Channel at 60 GHz for Vehicular Communication(IEEE, 2021-09); ; ; ; The use of millimeter wave communication gets more and more attention with the growth of 5G, particularly with regard to promising applications like platooning or intersection assistant. However, for the design of those applications, a clear understanding of the channel behaviour at these frequencies is mandatory. Aiming to fill the gap of missing channel models for frequencies above 60 GHz, this paper investigates the channel characteristics necessary for future vehicular applications. Based on the commercially available channel simulation tool WinProp, Altair, a typical vehicular scenario in the city center of Berlin (Germany) is analysed. The simulation results are verified with channel sounding measurements at 60 GHz. Hereby, the dependency of the channel with respect to scatterers like trucks and cars is evaluated. Reflection paths were found to be up to 30 dB higher in received power compared to scattering ones. Further, a platooning scene, consisting of 20 vehicles driving towards each other while communicating, is investigated for various time steps and receiver points. In this context, different antenna types are introduced to reduce interfering signals nearby the major transmitter, achieving a decrease in interference power of up to 20 dB.Publicationtype: Conference Paper58 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Integrated Classification and Localization of Targets Using Bayesian Framework In Automotive Radars(IEEE, 2021-06); ; ; ; Automatic radar based classification of automotive targets, such as pedestrians and cyclist, poses several challenges due to low inter-class variations among different classes and large intra-class variations. Further, different targets required to track in typical automotive scenario can have completely varying dynamics which gets challenging for tracker using conventional state vectors. Compared to state-of-the-art using independent classification and tracking, in this paper, we propose an integrated tracker and classifier leading to a novel Bayesian framework. The tracker’s state vector in the proposed framework not only includes the localization parameters of the targets but is also augmented with the targets’s feature embedding vector. In consequence, the tracker’s performance is optimized due to a better separability of the targets. Furthermore, the classifier’s performance is enhanced due to Bayesian formulation utilizing the temporal smoothing of classifier’s embedding vector.Publicationtype: Conference Paper113 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , A Bayesian framework for integrated deep metric learning and tracking of vulnerable road users using automotive radars(IEEE, 2021-05-05); ; ; ; With the recent advancements in radar systems, radar sensors offer a promising and effective perception of the surrounding. This includes target detection, classification and tracking. Compared to the state-of-the-art, where the state vector of classical tracker considers only localization parameters, this paper proposes an integrated Bayesian framework by augmenting state vector with feature embedding as appearance parameter together with localization parameter. In context of automotive vulnerable road users (VRUs) such as pedestrian and cyclist, the classical tracker poses multiple challenges to preserve the identity of the tracked target during partial or complete occlusion, due to low inter-class (pedestrian-cyclist) variations and strong similarity between intra-class (pedestrian-pedestrian). Subsequently, feature embedding corresponding to target's micro-Doppler signature are learned using novel Bayesian based deep metric learning approaches. The tracker's performance is optimized due to a better separability of the targets. At the same time, the classifiers' performance is enhanced due to Bayesian formulation utilizing the temporal smoothing of the classifier's embedding vector. In this work, we demonstrate the performance of the proposed Bayesian framework using several vulnerable user targets based on a 77 GHz automotive radar.Publicationtype: Journal ArticleScopus© Citations 16 146 248 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Channel Characterization at 77 GHz for Vehicular Communication(2020-12-16); ; ; ; With the rapid development of modern communication technologies, a well understanding of the channel is essential. Therefore, in this paper, the channel characteristic for mm Waves is investigated. Modern intelligent transportation systems require higher data rates and, therefore, more bandwidth. Thus, higher frequencies, which can provide these needs, get more and more in focus in the context of 5G, As part of our current research, the proposed work aims to give a deeper understanding of the channel characteristics and wave propagation at 77 GHz, especially for automotive use cases like platooning or intersection assistant. Typical traffic scenarios, starting with a rural street and proceeding with a highway scenario, were reconstructed using the commercial simulator WinProp. Furthermore, the scenarios were evaluated by including additional obstacles, like vehicles and traffic signs. Time-varying environments with driving vehicles with either line-of-sight (LOS) as well as non-LOS (NLOS) connections were investigated. The results reveal significant changes in the channel impulse response for one time step, when a new reflection path gets available. This is especially relevant in the NLOS case, since the scattering and reflection paths can have more than 25 dB difference in receiving amplitude. Overall, the antenna alignment and the environment had a strong influence. Thus, it could be shown that scatters like traffics signs cause power fading in the LOS connection, whereas they have benefits in the NLOS case, by enabling communication at all. In future research, the channel models will be further addressed and expanded to design a link that is capable of communicating safety-relevant information and hold the high demands on latency.Publicationtype: Conference Paper91