Ralph H. Rasshofer

Car2X-based perception in a high-level fusion architecture for cooperative perception systems

In cooperative perception systems, different vehicles share object data obtained by their local environment perception sensors, like radar or lidar, via wireless communication. In this paper, this so-called Car2X-based perception is modeled as a virtual sensor in order to integrate it into a highlevel sensor data fusion architecture. The spatial and temporal alignment of incoming data is a major issue in cooperative perception systems. Temporal alignment is done by predicting the received object data with a model-based approach. In this context, the CTRA (constant turn rate and acceleration) motion model is used for a three-dimensional prediction of the communication partners motion. Concerning the spatial alignment, two approaches to transform the received data, including the uncertainties, into the receiving vehicles local coordinate frame are compared. The approach using an unscented transformation is shown to be superior to the approach by linearizing the transformation function. Experimental results prove the accuracy and consistency of the virtual sensors output.

Circularly polarized millimeter-wave rectenna on silicon substrate

A circularly polarized (CP) silicon-integrated W-band rectenna (r_e_c_t_ifying ante_n_n_a_) for use in six-port polarimetric radar systems was designed, numerically optimized, and fabricated. A rigorous numerical optimization was performed by using the supergrid method (SGM). The rectenna applies a novel low-loss purely planar dual-patch antenna (DPA) layout, which allows the receiver to be manufactured using monolithic integration. The measurement results for the receiver demonstrate an excellent cross-polarization discrimination (XPD) >14 dB @ 76 GHz over a wide range of the scan angle (12 dB @ /spl plusmn/20/spl deg/).

Functional requirements of future automotive radar systems

Automotive radar systems are starting to divide into two groups: Highly specialized stand alone low-cost sensors targeting high volume markets and high-performance multi purpose sensors used in sophisticated data fusion architectures. While the first group focuses on ultimate cost reduction, the latter provides the basis for future high-performance driver assistance functions which might be deployed in middle-to luxury-class cars. This paper addresses both sensor groups and lists OEM requirements resulting from the specifications of future driver assistance functions and from upcoming perception system architectures. The second part of this contribution addresses recent trends in automotive radar sensing technology.

A direction sensitive, integrated, low cost Doppler radar sensor for automotive applications

We fabricated and tested an integrated, low cost, W-band Doppler radar sensor, capable to provide direction sensitive velocity information. The front-end consists of an active integrated antenna in self-mixing operation and a surface-wave coupled, mixing rectenna, providing full homodyne I/Q-detection. In the front-end, we employed only low cost silicon monolithic millimeter wave integrated circuits (SIMMWIC). Measured results show excellent performance of the sensor.

Cooperative Multi-User Detection and Localization for Pedestrian Protection

In the proposed cooperative sensor system pedestrians carry a reactive transceiver which is interrogated by a localization and tracking unit in the car. The prototype system applies Round-Trip Time-of-Flight (RTOF) echniques for the determination of the distance between the transponder and the demonstrator vehicle. A smart antenna array integrated into the car is used to determine the Direction-of-Arrival (DoA) of the transponders response signal. Knowing the distance and azimuth angle relative to the car, the pedestrians position and movement are calculated. These data are used as input for a highly reliable collision warning and collision mitigation system. The sensor system is capable of addressing a huge number of communication partners within each measurement cycle. Additionally secure burst identification is ensured for a robust localization and the suppression of unwanted co-channel interference. This is achieved by using pseudo random coded signals with a Time Division Multiple Access (TDMA) method. The distance accuracy was improved by introducing a new mirror technique in combination with an interpolation algorithm. The prototype localization system set up at 2.4GHz covers a range up to 200m in free field condition. With the current system a distance resolution down to cm and an angular measurement accuracy of about 1 degree have been achieved.

Correlation between subjective driver state measures and psychophysiological and vehicular data in simulated driving

Advanced driver assistance systems require better knowledge of the drivers state. This would allow for adapting driving support functions, e.g. adaptive automation. To detect the emotional and cognitive state of the driver, it is necessary to know which signals contain accurate information about the state. In this paper the results of a driving simulator study, in which different emotional and cognitive states were induced in 46 subjects via traffic scenarios, are presented. In the study, psychophysiological and vehicular data was measured in addition to subjective state estimations of the subjects. A correlation analysis confirmed that physiological data can potentially predict subjective driver states.

Automotive Radome Design - Reflection Reduction of Stratified Media

This letter summarizes the electromagnetic field theory of stratified media and provides an approach for designing antireflection structures. The theory of stratified media has been well known for over half a century. Nevertheless, the calculation of reflection degradation is often done using the assumption of load matching. This assumption turns out to be not accurate enough for efficient design of antireflection media for radomes in the millimeter-wave regime. We introduce an approach for optimized matching in this case and show a tolerance analysis of an example structure (rib profile).

Generic architecture for simulation of ADAS sensors

With rising complexity of advanced driver assistance systems (ADAS), development quality and speed are of increasing importance. Virtualization contributes towards this goal by providing a controlled environment even before hardware prototypes are available. One crucial component is a realistic description of the vehicles sensors by means of sensor models. We propose a generic, modular architecture for the design of such models. The cars surroundings are represented by a list of objects that is modified during the virtualized sensing process. The result can then provide stimulation of ADAS function algorithms. Using the proposed architecture as a framework, we construct an exemplary model employing a stochastic variation of target object positions.

Long-term stability of passive millimeterwave circuits on high-resistivity silicon substrates

We investigate the long-term performance of passive millimeterwave microstrip circuits on high resistivity silicon substrates. Three types of test structures were exposed to harsh environmental conditions such as thermal stress, humidity and organic vapors for 400 hours. Measured results at 38 GHz showed excellent long-term stability of the circuits. No significant difference between chips with and without SiO/sub 2/ passivation layer has been found though chips with passivation showed larger photosensitivity and higher loss.

An I-Q mixer at 76.5 GHz using flip-chip mounted silicon Schottky diodes

Silicon Schottky diodes show the advantages of low 1/f noise combined with low cost. We use these diodes in hybrid flip-chip configuration to build a novel I-Q mixer at 76.5 GHz for automotive applications. The realized mixer shows promising features such as 10 dB conversion loss @ 100 kHz IF, LO-to-RF isolation better than 25 dB, and average IF noise power of -80 dBm.