This paper describes optically semi-transparent and flexible microstrip interconnects and patch antennas at millimeter wave frequencies. They are realized on a 5-mil thick transparent PET film by patterning honeycomb shape metal grids using a standard lithographic process. Dimensions of the employed metal grids were optimized for low insertion loss and > 70% optical transmission at 77 GHz to enable integration into non-conventional surfaces for potential automotive radar applications. The fabricated microstrip lines demonstrate lowloss interconnects with < 0.35dB/mm loss at 77 GHz and ~73% optical transmission from 390 nm to 700 nm. The fabricated semi-transparent microstrip patch antennas show radiation characteristics comparable to that made of standard solid metal.

In this paper, we present a 122-GHz low-cost, robust, and small form-factor frequency modulated continuous wave (FMCW) radar sensor in package. The sensor is based on 130nm SiGe BiCMOS transceiver chip which is integrated in an embedded wafer level ball grid array (eWLB) package. The chip features a fully-differential, frequency-multiplier based bistatic radar with IQ receiver and IQ modulation in the transmitter. The highlight of the sensor includes a transmit and a receive antenna array designed in the redistribution layer of the eWLB package. Each antenna array comprise 43 rhombic elements providing a measured gain of around 15 dB. The radar system-in-package demonstrates a measured effective isotropic radiated power of 21.5dBm at 122 GHz which is the highest EIRP up to date for a D-band radar sensor not using an external lens. The radar sensor was also functionally tested in an FMCW configuration after mounting on an FPGA baseband board.

Today's cars are equipped with radar sensors, which provide precise information about the distance, speed and angle to surrounding objects on the road. This information is essential for modern driver assistance systems such as adaptive cruise control or brake assistance systems. Further, it enables future autonomous driving features. Most importantly, however, the accuracy and range of the radar sensors are critical for the safety of the car occupants as well as other daily road users. This is of particular significance since around 90 percent of all rear-end collisions with personal injuries occur due to human mistakes. Assuming all cars on the road to be equipped with emergency brake systems, up to 72 percent of these collisions could be prevented. For reasons of car appearance as well as protection of the device itself, the radar sensors are often mounted right behind the bumper. This, however, causes unwanted signal reflections from such. Particularly, the reflections yield so-called short-range (SR) leakage, which superimposes reflections of true objects that have to be detected most precisely. In fact, the SR leakage limits the achievable sensitivity and accuracy of the radar sensor severely. As a consequence, driver assistance systems may react delayed in critical situations. In this talk, novel concepts for SR leakage cancelation in automotive applications are presented. These concepts are the first known solutions of their kind that can be integrated holistically within a monolithic microwave integrated circuit (MMIC) operating at 77 GHz. The achievable performance gain is shown based on simulation as well as measurement results from the developed hardware prototype.

Cooperative Intelligent Transport Systems (C-ITS) promise to enhance the state-of-the-art in the mobility of people and goods by making transport cleaner, safer, and more efficient [2]. To this aim, all road users will be linked to each other and the road infrastructure to make traffic more efficient, cleaner, and safer. For example, Vehicle-To-Vehicle (V2V) and Vehicle-To-Infrastructure (V2I) communication enable coordination of vehicular traffic and advanced route management in transport networks [7]. Some C-ITS services require high data rates, but don't care much about latency, whereas others need to enforce strict demands on reliability and latency for short messages. These contrasting demands drive the diversification of radio access on board of vehicles. Challenges for C-ITS are posed by the nonstationary time-frequency-selective fading processes in vehicular channels. Fortunately, the nonstationary vehicular fading may be characterized by assuming local stationarity for a finite region in the time- frequency plane. For such region, the wide-sense stationarity and uncorrelated scattering assumptions hold approximately. Thus, it makes sense to characterize the channel by a local scattering function (LSF). Estimates for the LSF from measurements collected in the DRIVEWAY'09 campaign at 5-6 GHz are discussed focusing on ITS scenarios. Subsequently, the time- frequency-varying power delay profile (PDP) and the time-frequency-varying Doppler power spectral density (DSD) are discussed. Based on these, the time-frequency-varying delay and Doppler spreads are evaluated. High delay spreads are observed in situations with rich scattering, whereas high Doppler spreads characterize drive-by scenarios. Early LSF results for 59.75-60.25 GHz millimeter wave V2V channel measurements in an urban street (Vienna, Austria) are presented. Measurements have been acquired in September 2017 with a time-do- main channel sounder. Estimates for delay and Doppler profiles are evaluated from the LSF for overtaking vehicles. Passenger cars are associated with a single Doppler trajectory, whereas lorries show up in the LSF with multiple Doppler trajectories. These propagation-related channel characteristics translate into packet transmission error sequences exhibiting strong temporal dependencies. Finally, we discuss packet error models of low complexity for large-scale C-ITS emulation.

It is possible to transmit ETSI ITS-G5 standard compliant packets using commercial off-the-shelf Wi-Fi cards based on the Atheros 9k chipset under Linux operating system. Especially for road safety applications, low latency transmission of critical messages is required. We measure latencies in such a system in order to identify causes of delays and therefore areas for improvement. First analysis using a system without any optimizations revealed transmission and reception delays, the causes of which are to be investigated further.

The development of mobility is moving rapidly towards intelligent connected vehicles, further advances in automation of driving functions and eventually self-driving vehicles. In future automobiles, wireless systems for communications and sensing will play an even more crucial role than today. Within automobiles as well as between multiple automobiles, car-to-car, (C2C) and the surrounding road infrastructure, car-to-infrastructure (C2I), reliable wireless links become indispensable. The communication standards, the vehicular electronic control systems, as well as the hybridisation of cellular and ad-hoc networks are moving towards increased complexity. Additionally, the performance validation is moving towards Over-The-Air (OTA) system testing. Consequently, the requirements on test facilities, equipment and methods increase accordingly

In this paper, we give an overview of state-of-the art full-vehicular testing with emphasis on recent developments in advanced post-processing techniques to evaluate final system performance. The discussion is supported by measured examples on a car mock-up.

In this paper, we present a concept for wireless local positioning with multiple backscatter transponders based on the switched-injection locked oscillator (SILO). It allows reliable localization and ego-motion estimation of autonomous robots in indoor and outdoor environments. We derive the range-doppler spectrum of the beat signal that is induced by a single transponder in a FMCW radar system. With multiple transponders, a localization network is created that allows multilateral and multiangulation based positioning. For an initial evaluation, we built a miniaturized transponder for the 24\ GHz ISM band and proofed the concept with a straightforward measurement, including two transponders, a moving ground robot equipped with a FMCW radar system and a laser tracker reference.

This paper introduces an improvement of the target separation capability for radar systems. The separation capability depends on the used bandwidth and on the radar cross section (RCS) of the targets to be separated. In worst case scenarios only targets with a high range difference can be separated. In multi target scenarios the received signal is a superposition of all reflected signals. By using different starting frequencies, the superposition behavior is changed. The new algorithm uses this information to separate different targets. Simulations and measurements are used to validate the new method. With this algorithm it is possible to separate targets better than with a conventional FFT spectrum analysis.

Current automotive radar sensors enhance the angular resolution using a multiple-input multiple-output approach. The often applied time-division multiplexing scheme has the drawback of a reduced unambiguous Doppler velocity proportional to the number of transmitters. In this paper, a signal processing scheme is proposed to regain the same unambiguous Doppler as in the single-input multiple-output case with only one transmit antenna. Simulation and measurement results are shown to prove that the signal processing leads to an enhanced unambiguous Doppler velocity estimation.

This paper highlights the parameterization limitations of MIMO OFDM radar especially in multi-user scenarios. First the parameter dependencies are depicted and the resulting limitations are derived. Based on this it is demonstrated how the basic scheme can be modified to generate orthogonal channels, either multiple transmitters of the same system or for different users to operate in an almost interference free state. The focus is on automotive applications for highly automated driving (HAD) that require both, high range and velocity resolution as well as fail save operation in scenarios with multiple users. Additionally, the limitations of Compressed Sensing in combination with OFDM radar to overcome some of the previous limitations are analyzed. Analytical studies as well as simulations have been done to depict the mentioned limitations.

Ground Penetrating Radar (GPR) is one of the tools supporting mine detection. In this contribution a wide-band frequency modulated continuous wave (FMCW) GPR from 1 GHz to 4 GHz in a bistatic configuration is presented. This radar is designed so that it can be mounted on an unmanned aircraft vehicle (UAV). A compromise between weight, size, power consumption and penetration depth is found. The realization of the radar by means of frequency band splitting is presented. The merging of the two frequency bands is evaluated by measurements. The radar has been successfully integrated on a UAV and first measurements over a test field from the flight are presented.

Radar sensors are utilized for detection and classification purposes in various applications. In order to use deep learning techniques, lots of training data are required. Accordingly, lots of measurements and labelling tasks are then needed. For the purpose of pre-training or examining first ideas before bringing them into reality, synthetic radar data are of great help. In this paper, a workflow for automatically generating radar data of human gestures is presented, starting with creating the desired animations until synthesizing radar data and getting the final required dataset. The dataset could then be used for training deep learning models. A classification scenario applying this workflow is also introduced.

We present an approach to detect and locate non-cooperative UAVs from their micro-Doppler signature using a narrowband radar in a multi-site configuration. We describe a method for the localization of rotating objects with the geometric information obtained exclusively from their micro-Doppler signatures. This approach only requires very simple transceivers with CW waveforms, in a cost-effective multi-site architecture. A convolutional neural network is used to detect and identify the UAVs by extracting the characteristic features of their micro-Doppler signature. We present simulated and preliminary experimental data that show the technical viability of this concept.

A novel approach for anti-personnel landmine detection using an unmanned aerial vehicle (UAV) in combination with a ground penetrating synthetic aperture radar (GPSAR) is presented. The objective of the system is to accelerate the process of land release in humanitarian demining. Suspicious objects shall be detected by the radar and marked for further investigations using different sensor principles.

The ground penetrating radar (GPR) module consists of a 1 GHz to 4GHz side-looking frequency modulated continuous wave (FMCW) radar and a real time kinematic global navigation satellite system (RTK GNSS). The image processing is done offline using a back-projection algorithm. In the theoretical part of this paper the system partitioning, the sensor module, and the position accuracy requirements are briefly described. First synthetic aperture radar (SAR) measurements are presented in the experimental part of this paper.

Global navigation satellite systems (GNSS) such as GPS are the de-facto standard for absolute and reliable unmanned aerial vehicles (UAV) localization due to their global availability and the maturity and wide spread of the underlying technology. However, as the spread and reliance upon autonomous UAV systems increases, the risk of jamming becomes a substantial safety problem. The area where jamming attacks are most likely to occur is the close vicinity of the UAV landing station. Hence, to ensure reliable operations, an alternative localization technology for the approach and landing phase is needed. This paper presents a localization system for 3D position estimation based on 24 GHz FMCW cooperative radar. It has very low infrastructure requirements thanks to the combination of ranging and direction-of-arrival (DOA) estimation techniques employed in the ground station. Jamming resistance is improved due to the large bandwidth of 250 MHz. A measurement campaign was conducted to prove the feasibility of the proposed concept.

DENSO is one of global ADAS supplier and has been mass-producing millimeter-wave electronically scanning radar for automotive use since 2003, which is the longest market track record in the world. In my presentation, I will introduce these development histories and current activity for the next generation ADAS. Contents

1. ADAS trend and Radio act in JAPAN
2. Introduction of DENSO's ADAS activity
3. DENSO's radar technologies /Planar Antenna /C-MOS MMIC /High resolution Signal Processing (Null Scanning method) /UWB(Ultra Wide Band)

Automated Driving is expected to contribute to various factors that affect our daily life, while brings new business and technological challenges for various stakeholders involved. With the advent of automated driving functions, especially with the broad availability of vehicles that will be capable of supporting higher automation levels, the need for synchronization and coordination among vehicles becomes increasingly necessary. Vehicles will be connected to the Internet and will communicate directly with each other to extend their perception beyond the capabilities and the range offered by their integrated sensors (camera, radar, ultra-sound range finders, etc.). The ability to exchange related information (e.g., cooperative awareness, road hazards, etc.) is expected to improve the decision making process for self-driving. To enable this, current Intelligent Transport System (ITS) services make use of communication links in the form of: a) Vehicle-to-Vehicle (V2V), to exchange information directly between vehicles, b) Vehicle-to-Infrastructure (V2I), to exchange data between a vehicle and road infrastructure (e.g., traffic lights), c) Vehicle-to-Pedestrian (V2P), to exchange information between vehicles and mobile devices carried by a pedestrian, a cyclist, etc, d) Vehicle-to-Network (V2N), to exchange information between a vehicle and a backend server in the network and to get additional services like map updates, fleet-based data collection, etc. Therefore, the summarizing term Vehicle-to-Everything (V2X) describes any communication involving a vehicle as a source or destination of a message. Cooperative lane change, cooperative collision avoidance, and platooning are typical examples of V2X use cases, where connected automated vehicles participate. The involved vehicles trigger a specific use case for safety reasons (e.g., emergency maneuver) or for efficient traffic flow (e.g., platooning) using the monitoring data from the installed sensors, together with the information received from neighboring vehicles. Thereinafter, the automated vehicles undertake to coordinate and plan their maneuvers or trajectories in order to address the triggering event, based on the built environmental perception. As vehicles advance towards higher automation levels and need to deal with increasingly complex road situations, there will be a need for a complementary communication technology for the exchange of cooperative information with higher bandwidth and improved reliability. In connected automated vehicles the performance requirements are more stringent, with certain use cases requiring very reliable communication links (>99.99%), with much lower maximum end-to-end latency (1-10 ms), and higher data rate. Vehicles will be also connected to one or more intelligent transportation systems (ITS) application servers (e.g., for traffic management services) via V2N, which does not require strict latency or reliability requirements. However, there is a specific use case, where numerous technical hurdles need to be overcome, especially for the cellular technologies: vehicle tele-operation on public roads. Tele-operation refers to vehicles that are controlled over the network by a remote operator. With the aid of sensors mounted on the vehicle, a remote operator can control one or multiple concurrent vehicles. With the promise of high reliability, availability, and sub-10 ms end-to-end delays, 5G systems have the potential to enable tele-operated vehicles. Furthermore, there are also use cases where different communication modes and technologies (V2V, V2I, sensors, ITS application) could be combined. 5G has the benefits for enabling C2C communication that will address the stringent performance requirements (delay, reliability, capacity, etc.) that autonomous driving sets for the communication layer. 5G technological innovation is expected to enable robust and resilient communication among vehicles and other entities by using multiple communication links (direct/sidelink, cellular, etc.) with guaranteed QoS, where changes in the QoS availability could be notified to the vehicles a-priori. Multi-antenna and diversity techniques including MIMO are crucial enablers for achieving high transmission reliabilities under latency constraints and higher data rates. Efficient resource management algorithms on New Radio (NR) sidelink and cellular interfaces, and 5G radio-assisted positioning are some of the 5G technologies that could bring substantial improvements to the performance of the communication layer to address the requirements of vehicular communications, operating in different frequency bands (below and above 6 GHz). 5G will provide a flexible functional architecture, based on different business considerations. Network slices are means to realize novel business models, enabling differentiated management of network resources and traffic handling according to the varying characteristics of V2X services.

An essential feature of the signals of global satellite navigation systems (Global Satellite Navigation Sys- tems, GNSS) is their low spectral power density at the location of the GNSS navigation receivers of the users. The power of the GNSS signals recorded by a typical user antenna is approximately 25 dB below the power of the thermal noise in the relevant frequency band. Only the spread-spectrum structure of GNSS signals allows GNSS navigation receivers to determine position and time information by exploiting the processing gain that can be obtained through correlation reception. One of the most significant sources of error in satellite navigation is terrestrial radio interference (Radio Frequency Interference, RFI) which typically can be several orders of magnitude more powerful than the useful GNSS signals. Such interfering signals are usually caused unintentionally by imperfections of technical systems or inadequate technical compatibility when frequency bands are used jointly. Unfortunately, there is currently also an increasing tendency to be observed of deliberately and illegally emitted interference signals e.g. by personal private devices. All of the interference signals mentioned above have the potential to interfere with the reception of GNSS signals in navigation receivers. The impairments range from faulty position determination - and thus a threat to the integrity of the determined position - to the complete prevention of position determination - and thus a restriction of the availability of GNSS. Even more critical are spoofing type of deception attempts, which can also appear intentionally or unintentionally. The underlying danger of this type of attack is the presence of duplicates of the original GNSS signal, so that the receiver cannot distinguish between original and forgery. In a successful spoof- ing attack, the GNSS receiver uses the counterfeit signals, leading to incorrect positioning and/or timing. Particularly for safety-critical applications that require reliable and robust position information, such as automatic landing of aircraft, autonomous driving etc., the abovementioned impairments are an unacceptable hazard. Therefore, it is imperative to develop countermeasures that ensure a high robustness of GNSS navigation receivers against deliberate and unconscious interference signals. The keynote will provide an overview of the most important origins of unintentional and intentional interfering signals in the satellite navigation environment as well as current approaches and methods for sup- pressing their impact on GNSS receivers. The presentation of a robust GNSS receiver demonstrator, which implements selected approaches based on FPGA technology, as well as performance evaluations and validations in the real world, complete the presentation.

In this paper, we present a concept for wireless local positioning with multiple backscatter transponders based on the switched-injection locked oscillator (SILO). It allows reliable localization and ego-motion estimation of autonomous robots in indoor and outdoor environments. We derive the range-doppler spectrum of the beat signal that is induced by a single transponder in a FMCW radar system. With multiple transponders, a localization network is created that allows multilateral and multiangulation based positioning. For an initial evaluation, we built a miniaturized transponder for the 24\ GHz ISM band and proofed the concept with a straightforward measurement, including two transponders, a moving ground robot equipped with a FMCW radar system and a laser tracker reference.

This paper introduces an improvement of the target separation capability for radar systems. The separation capability depends on the used bandwidth and on the radar cross section (RCS) of the targets to be separated. In worst case scenarios only targets with a high range difference can be separated. In multi target scenarios the received signal is a superposition of all reflected signals. By using different starting frequencies, the superposition behavior is changed. The new algorithm uses this information to separate different targets. Simulations and measurements are used to validate the new method. With this algorithm it is possible to separate targets better than with a conventional FFT spectrum analysis.

Current automotive radar sensors enhance the angular resolution using a multiple-input multiple-output approach. The often applied time-division multiplexing scheme has the drawback of a reduced unambiguous Doppler velocity proportional to the number of transmitters. In this paper, a signal processing scheme is proposed to regain the same unambiguous Doppler as in the single-input multiple-output case with only one transmit antenna. Simulation and measurement results are shown to prove that the signal processing leads to an enhanced unambiguous Doppler velocity estimation.

This paper highlights the parameterization limitations of MIMO OFDM radar especially in multi-user scenarios. First the parameter dependencies are depicted and the resulting limitations are derived. Based on this it is demonstrated how the basic scheme can be modified to generate orthogonal channels, either multiple transmitters of the same system or for different users to operate in an almost interference free state. The focus is on automotive applications for highly automated driving (HAD) that require both, high range and velocity resolution as well as fail save operation in scenarios with multiple users. Additionally, the limitations of Compressed Sensing in combination with OFDM radar to overcome some of the previous limitations are analyzed. Analytical studies as well as simulations have been done to depict the mentioned limitations.

Ground Penetrating Radar (GPR) is one of the tools supporting mine detection. In this contribution a wide-band frequency modulated continuous wave (FMCW) GPR from 1 GHz to 4 GHz in a bistatic configuration is presented. This radar is designed so that it can be mounted on an unmanned aircraft vehicle (UAV). A compromise between weight, size, power consumption and penetration depth is found. The realization of the radar by means of frequency band splitting is presented. The merging of the two frequency bands is evaluated by measurements. The radar has been successfully integrated on a UAV and first measurements over a test field from the flight are presented.

Radar sensors are utilized for detection and classification purposes in various applications. In order to use deep learning techniques, lots of training data are required. Accordingly, lots of measurements and labelling tasks are then needed. For the purpose of pre-training or examining first ideas before bringing them into reality, synthetic radar data are of great help. In this paper, a workflow for automatically generating radar data of human gestures is presented, starting with creating the desired animations until synthesizing radar data and getting the final required dataset. The dataset could then be used for training deep learning models. A classification scenario applying this workflow is also introduced.

We present an approach to detect and locate non-cooperative UAVs from their micro-Doppler signature using a narrowband radar in a multi-site configuration. We describe a method for the localization of rotating objects with the geometric information obtained exclusively from their micro-Doppler signatures. This approach only requires very simple transceivers with CW waveforms, in a cost-effective multi-site architecture. A convolutional neural network is used to detect and identify the UAVs by extracting the characteristic features of their micro-Doppler signature. We present simulated and preliminary experimental data that show the technical viability of this concept.

A novel approach for anti-personnel landmine detection using an unmanned aerial vehicle (UAV) in combination with a ground penetrating synthetic aperture radar (GPSAR) is presented. The objective of the system is to accelerate the process of land release in humanitarian demining. Suspicious objects shall be detected by the radar and marked for further investigations using different sensor principles.

The ground penetrating radar (GPR) module consists of a 1 GHz to 4GHz side-looking frequency modulated continuous wave (FMCW) radar and a real time kinematic global navigation satellite system (RTK GNSS). The image processing is done offline using a back-projection algorithm. In the theoretical part of this paper the system partitioning, the sensor module, and the position accuracy requirements are briefly described. First synthetic aperture radar (SAR) measurements are presented in the experimental part of this paper.

Global navigation satellite systems (GNSS) such as GPS are the de-facto standard for absolute and reliable unmanned aerial vehicles (UAV) localization due to their global availability and the maturity and wide spread of the underlying technology. However, as the spread and reliance upon autonomous UAV systems increases, the risk of jamming becomes a substantial safety problem. The area where jamming attacks are most likely to occur is the close vicinity of the UAV landing station. Hence, to ensure reliable operations, an alternative localization technology for the approach and landing phase is needed. This paper presents a localization system for 3D position estimation based on 24 GHz FMCW cooperative radar. It has very low infrastructure requirements thanks to the combination of ranging and direction-of-arrival (DOA) estimation techniques employed in the ground station. Jamming resistance is improved due to the large bandwidth of 250 MHz. A measurement campaign was conducted to prove the feasibility of the proposed concept.

Environment models are necessary for autonomous driving. The distinction between drivable and non-drivable underground is elementary. This paper presents adaptions for radar based occupancy gridmaps, which are a common representation of the environment. In contrast to standard occupancy gridmaps or in general standard inverse radar sensor models, our approach works with velocity dependent parameters and extends free space calculations. Consequently, the map quality varies less and the information content of the ego vehicle's immediate vicinity is higher. Experiments with ground truth data show that the proposed algorithm produces accurate environment models in urban scenes.

The motion of a car has a vast impact on the relative measurements of all environmental sensors. Consequently, the precise knowledge of the ego-motion of the sensor equipped car is important for higher level signal processing like mapping or the estimation of the target parameters. This can be achieved by multiple simple radar sensors with a single measurement. The range and the velocity measurements conducted by multiple sensors that are observing a common field of view are jointly processed and filtered for stationary targets. This dataset is used to determine the vehicles velocity and yaw-rate in the x- and y- plane. An advantage of the proposed approach is the simplicity of the sensors which do not need the capability of angle estimation in the azimuth plane, while still providing comparable results. In addition, this approach works with a minimum number of two sensors.

Prediction of the object movement from sensor data in the automotive sector is a widespread research and development topic. Dependent on the used sensor types, object tracking has been established over several measurement cycles. A prominent example of this is the Kalman filter. In time critical scenarios with less reaction time tracking over a number of measurement cycles is not suitable. To detect object movement within one single measurement cycle only the radar sensor is a candidate, due to the ability to measure the velocity of objects instantaneously by using the Doppler effect. A new approach to estimate the direction of movement of cyclists within one measurement cycle is introduced and explained in this paper. It is based on the approximation of the shape from a cyclist. The approximation is performed with two different methods and the solutions are compared with each other. To validate the results of the direction estimation, simulated and measured radar data are exercised.

Lane Change Assist, Lane Departure Warning, Lane Keeping System, and Automated Driving all require the sensors in vehicles to be able to detect the road and lane boundaries reliably and accurately under all circumstances. However, such functions rely currently solely on camera sensors, which in result unavoidably affects the stability of the whole system under certain extreme lighting conditions. Therefore, this paper shows a verification test result of a novel approach in lane boundary detection with an automotive radar and a specific kind of reflectors. A new approach of road course prediction and vehicle localization by utilizing its Range-Doppler characteristics is also presented in this paper.

This paper deals with gesture recognition using a 77 GHz FMCW radar system based on the micro-Doppler (μD) signatures. In addition to the Doppler information, the range information is also available in the FMCW radar. Therefore, it is utilized to filter out the irrelevant targets. We have proposed five micro-Doppler based handcrafted features for gesture recognition. Finally, a simple k-nearest neighbor (k-NN) classifier is applied to evaluate the importance of the five features. The classification results demonstrate that the proposed features can guarantee a promising recognition accuracy.

Fifth generation (5G) mobile communication promises many new vertical service areas beyond communication and internet data access. We propose CPCL (Cooperative Passive Coherent Location) being a distributed MIMO radar service which could be offered by mobile radio network operators as a service for public user groups. CPCL comes as an inherent part of the radio network and takes advantage from the most important key features proposed for 5G. It extends the well-known idea of passive radar (also known as Passive Coherent Location, PCL) by introducing cooperative principles. These range from co- operative, synchronous radio signaling and MAC up to radar data fusion on sensor and scenario levels. Using the software defined radio and network paradigms and real-time mobile computing facilities, CPCL promises to become a ubiquitous radar service which may be adaptive, reconfigurable, and, hence, cognitive. Because CPCL makes double use of radio resources (both, in terms of frequency bands and hardware), it can be considered a green technology. Although we introduce the CPCL idea from the viewpoint of vehicle-to-vehicle/infrastructure (V2X) communication, it can be definitely also applied for many other applications in industry, transport, logistics, and for safety and security applications. As CPCL, first of all, is a radio technology we will put focus on multipath radar signal processing and PHY resource exploitation.

It is quite remarkable that so little attention has been paid so far to develop NLOS path loss models for V2V communication in urban intersections, since these are one of the most safety critical scenarios for V2V safety applications. The Mangel model [1] is an exception, providing an NLOS model based on extensive measurements, and is by far the most used one today. However, there are limitations with the Mangel model with respect to reciprocity and that the model does not consider other vehicles obstructing the NLOS communication link. These two limitations have been considered in the analysis in this paper of V2V measurements between six vehicles in the city of Gothenburg, Sweden. We propose a measurement based reciprocal V2V NLOS path loss model, for intersections, comprising single interactions from the intersection center area together with multiple interactions from the building walls. When the communication link is obstructed by other vehicles, the contribution from the single interactions is insignificant. On the other hand, when both the communicating vehicles have LOS towards the intersection, it is the single interactions factor that dominates at large distances. The channel gain in our measurements is generally higher compared to what is predicted with the Mangel model. With our proposed model the communication links will have longer ranges compared to the cases using the Mangel model, another aspect is that interference levels from other vehicles will increase. We have also presented measured packet success ratios, PSRs, for 25 simultaneous communication links in the intersections. The results show that the probability to receive information from another vehicle around the corner approaching the same intersection is quite high before the moment when it is too late to take actions to avoid a collision. By using V2V communication as a complement to the on-board sensors (e.g., radars, cameras, and lidars), the traffic safety can be increased in intersections, even during dense traffic when the vehicles are shadowing each other. Our proposed path loss model needs to be validated in more intersections. We will present results regarding the auto- and cross-correlation of the multilink shadowing process as future work.

For critical vehicular communication services, such as traffic safety and traffic efficiency, it is advisable to design systems with robustness as the main criteria, possibly at the price of reduced peak performance and efficiency. We describe a simple, low-cost method for combining the output of L nonideal (i.e., nonisotropic) antennas to the input signal to a single-port receiver with the aim to guarantee robustness, i.e., to minimize the probability that K consecutive packets arriving from the worst-case angle-of-arrival are decoded incorrectly. To minimize complexity, the combining network does not estimate or use channel state information (complex channel gains, noise levels, etc.). The combining network consists of L − 1 analog phase shifters whose phases are affine functions of time. For a general L and the case when the packet error probability decays exponentially with the received SNR, the optimum slopes of the affine functions can be computed by solving an optimization problem that depends on the antenna far field functions. We provide an analytical solution for the special case of L = 2 antennas, which turns out to be independent of the antenna patterns. In an experimental setup consisting of two monopole antennas mounted on the roof of a Volvo XC90, the proposed combining method is shown to give significant performance gains compared to using just one of the antennas.

An automotive antenna set of four collinear arrays between 24.25 GHz and 29.5 GHz is introduced for application of 5G mobile communication in roof radomes of future cars. Each of the single antenna elements combines high gain in horizontal direction with an omnidirectional pattern which is well suited for application in multiple input multiple output (MIMO) transmission. In a phased array application the four antenna elements yield in the horizontal plane a high gain of 16 dBi at driving direction and 10.5 dBi in lateral direction which yields good access along the driving path. The gain of the antenna elements, efficiency, mutual coupling and the variable radiation characteristics of the antenna array are investigated by means of simulation and measurement.

A two-dimensional Rotman-lens feeding circuit was designed for multi-beam antenna using substrate integrated waveguide in multi-layer substrate in the millimeter-wave band. Metal patterns between the substrates are removed to form Rotman-lens in the E-plane as well as in the ordinary H-plane Rotman-lens feeding circuit. A complete two-dimensional multi-beam forming circuit is integrated in the multi-layer substrate. The simulated beam-forming performance was evaluated in this work.

Automotive radars are a keystone in the success of autonomous driving capabilities and many of modern ADAS functions witnessed today on highways and soon in urban areas. Their operation physics allows exceptional robustness in mostly all-weather conditions, and the unique access to Doppler information in real-time is remarkable. With the extended complexity and intelligence in these sensors, their testing is becoming even more challenging. In addition, sensor fusion combined to machine learning algorithms makes the testing conditions even more comprehensive and time consuming. Test drives are no more a cost-effective solution to satisfy these demands. Reproducible testing under controlled environment is therefore becoming vital to judge and verify vehicle responses in critical drive scenarios. In this talk, an overview of the radar testing methodologies and a visionary extension to achieve scene simulation for radar sensors will be delivered. Moreover, the state-of-the-art in testing automotive radars in motion is presented along with an exploration of a demonstration available during the ICMIM exhibition.

Automotive radar at 77 GHz is a key technology for the implementation of autonomous driving. The sensors and the systems associated with it are traditionally tested using real test drives that span several millions of kilometres. However, these test drives do not provide a reliable, reproducible nor controlled environment. Moreover, the costs and time effort associated with them is tremendous. Hence, novel test systems are required where experiments can be performed in an accurate, reliable, reproducible, and controllable way, by emulating a representative and realistic virtual test environment. The objective of this paper is to present an architecture and approach to augment real field tests. We use a hardware-in-the-loop approach that emulates a virtual radar environment corresponding to the scenarios defined for test and validation, and simulates the response behaviour of the interconnected automotive subsystems. The whole system will be installed within the Virtual Road Simulation and Test Area at the Technische Universit¨at Ilmenau. The solution comprises a holistic evaluation system where the radar-under-test is installed in the car, stimulated in a virtual electromagnetic environment, and its performance is evaluated in real time. The approach is to divide the system design problem into blocks of independent subsystems and their components.We analyse the critical parameters of each one of them, in order to arrive at a coherent system concept through interconnection of these sub-systems. We also discuss some of the initial aspects of our implementation.

Future autonomous automobiles will require extremely reliable radar sensors to guarantee the safety of road users. This requires test environments to verify that automotive radars work properly even in complex traffic situations. Radar target simulators can generate virtual scattering centers in front of the radar sensor and are therefore a promising verification technique. However, currently available systems only offer a very basic approach to manipulating scattering centers. With the workflow presented is this publication it becomes possible to decompose high-level descriptions of traffic scenarios into parameters required for radar target simulation. The workflow can therefore be used to generate realistic traffic scenarios and thus increases the reliability of automotive radar sensor tests.

Automotive radar systems become more and more indispensable for advanced driving assistance systems. Beside existing monostatic radar, bistatic radar sensing, like passive coherent location, provide additional options to augment the radar visibility of vulnerable road users. Assured by the planned coexistence of different wireless standards like ITS-G5 and LTEV, multiple new illuminators-of-opportunity can be applied to increase awareness and safety in complex traffic scenarios. Regarding the bistatic radar cross section of vulnerable road users, the frequency range of interest extends from about 450MHz to 6 GHz. This paper provides simulation and measurement approaches of bistatic radar cross sections of vulnerable road users like bicycles or pedestrians. Electromagnetic simulations and bistatic measurements of a bicycle are presented and compared. The results show reasonable agreement between simulation and measurement and provide new insight into the wireless environment in a frequency range rarely considered for radar sensing until to-date.

This paper presents the first measurement results of a new Orthogonal Frequency Division Multiplexing (OFDM) Multiple Input Multiple Output (MIMO) Radar prototype. It is designed to be operated in a dynamic multi-user environment where the individual sensor's RF bandwidth is very limited. Here, the radar uses only 30 MHz. The targeted applications are in the field of Urban (Air) Mobility (UAM) with a medium range coverage up to 250 m. Another requirement is to observe a wide field of view in azimuth of about 100 degree with a high resolution in azimuth. This was implemented by using a 2 x 8 MIMO processing. A 20 ms frame based evaluation of the OFDM symbols allows to obtain a very high velocity resolution of below 1 m/s. This paper explains details about the radar system, the waveform and the processing but the main part of the paper focuses on the evaluation of a dynamic measurement.

In this paper, we present a neural network based classification algorithm for the discrimination of moving from stationary targets in the sight of an automotive radar sensor. Compared to existing algorithms, the proposed algorithm can take into account multiple local radar targets instead of performing classification inference on each target individually.

Autonomously operating vehicles have a great potential to change multiple areas of daily life and technology in the very near future as they could offer lower running costs and higher reliability than human labor. While such systems are not yet quite capable to cope with complex and dynamic environments such as urban areas, autonomous vehicles are on the rise in certain simpler scenarios typical for the service and manufacturing sectors. A key prerequisite for autonomous operation in this context is reliable and accurate position estimation. While global navigation satellite systems (GNSS) such as GPS, resp. DGPS are widely used in outdoor applications, they fail to provide accurate and reliable localization results in many situations where the satellite signals are obstructed. These include areas near buildings, under cranes, and most indoor spaces. Wireless local positioning systems (WLPS) offer a way to make operation in such scenarios possible. In this talk a well-established WLPS for straddle carrier tracking in the container terminal logistics, as well as a very recent system for service robot localization in a hospital setting will be presented. The focus will be on the techniques used, the challenges faced during the development, and the results obtained in real-life scenarios.

The location awareness of objects is more and more important to mostly any kind of applications. Cars need to know where they are to drive autonomously, tools used in production are restricted to specific areas of operations and also sport coaches analyze and optimize the player's behavior due to their motion capabilities. Therefore, tracking systems of different kinds, like radio based, optical or laser based systems are designed for special purposes and applications. Datasheets are often not comparable due to different and not standardized testing methods of vendors as well as a hard forecasting of the performance of the system in a specific environment. To predict the performance without disturbing for example the production process, pseudo-real environment measurements will help to prevent expensive test installations on the user side.

Automotive radars operate mainly in 77 GHz and 79 GHz frequency bands as 24 GHz radars are phasing out. These radars are integrated into bumpers of the car or mounted behind so called design emblems. These plastic materials need to be RF transparent to let the signals from the radar sensors pass without distortion. However, due to different paintings and coatings of bumpers low RF transparency and non-homogeneity can cause radars to malfunction and lower the radar performance. Testing the RF performance of materials requires typically a network analyzer and takes a long time. This paper presents a new measurement method to test and verify radome materials within seconds.

As the number of radio systems in modern vehicles increases, electromagnetic interferences between them become an issue. In-vestigating and testing new wireless devices regarding the im-munity against radio disturbances and the emission of unwanted signals is essential to reach high reliability in later applications. Usually, these tests are carried out in fully anechoic chambers, to achieve defined, undisturbed wave propagation mimicking free-space conditions. Because the degree of achieving nearly ideal conditions can strongly influence measurement results, interna-tional standards provide procedures to validate test facilities. This paper discusses the influence of an electrically large antenna measurement arch on the norm compliance of the Virtual Road Simulation and Test Area (VISTA) of the Thuringian Center of Innovation in Mobility at the Technische Universität Ilmenau, with focus on the frequencies of modern wireless communication systems. Time domain information of a received test signal is used to identify the reflections and their sources in the chamber. This approach provides access to the site voltage standing wave ratio (sVSWR) demanded by standards and presents a novel general-ized technique for reliable site qualification in consideration of installed structures.

In a simulation for stimulation of automotive radar an analytic connection is established between SBR-Phong and radar cross section (RCS) computations. Phong is adapted and normed for this purpose. RCS values are computed for a primitive- (sphere) and a realistic geometry (car) using the adapted Phong approach, and are compared against computations from a commercial field simulation software.

The number of antennas in cars is increasing by the day. The interaction of an automotive antenna with the car chassis presents a prominent design challenge. While patch antennas are resilient to the underlying chassis metal, they suffer from poor impedance matching bandwidth for a low-profile design. Dipole antennas provide higher bandwidths, but the radiation performance degrades significantly if the dipole is mounted less than a quarter-wavelength apart from the car chassis metal. This paper shows through consistent simulation and measurement results the significant benefits of using a dipole over a high impedance surface (HIS) when low-profile antenna designs for confined mounting locations close to the chassis metal are required. It was observed that the HIS-dipole delivered 35% total efficiency for a mounting location that was only 3 mm apart from the chassis metal. At the same time, a well-defined resonance with a sufficient bandwidth and a positive peak realized gain (2.6 dBi) for the LTE-2600 frequency band (2500...2700 MHz) were obtained. These results indicate a stable and reliable performance of plastic embedded HIS-dipoles in a car, and open new avenues for novel close-to-chassis automotive antenna integration locations.

The antenna is an important part of the automotive radar system and influences the sensor's performance, especially its angular behaviour. One possibility is the use of a conformal antenna array to reach a larger view angle. However, the Rayleigh criterion cannot give an appropriate measure for the resolution because the array elements have different view angles. In this paper the incoherent superposition of two targets is assumed according to the Rayleigh criterion. In the first step the separability of two targets is checked with the patterns of the antenna array for all target angle combinations. Coming from this, an assumption allows the use of the ambiguity function to estimate the resolution. This works well for the main beam and gives the advantage of a fast calculation.

We propose a design strategy for optimizing antenna positions in linear arrays for far-field Direction of Arrival (DoA) estimation of narrow-band sources in collocated MIMO radar. Our methodology allows to consider any spatial constraints and number of antennas, using as optimization function the Weiss- Weinstein bound formulated for an observation model with random target phase and known SNR, over a pre-determined Field-of-View (FoV). Optimized arrays are calculated for the typical case of a 77GHz MIMO radar of 3Tx and 4Rx channels. Simulations demonstrate a performance improvement of the proposed arrays compared to the corresponding uniform and minimum redundancy arrays, even if the latter have a larger aperture.

Radar sensor has been an integral part of safety critical applications in automotive industry owing to its weather and lighting independence. The advances in radar hardware technology has made it possible to reliably detect objects using radar. The highly accurate radar sensors give multiple radar detections per object. In this work, a postprocessing architecture is presented which is used to cluster and track multiple detections from one object in practical multiple object scenarios. Furthermore, the framework is tested and validated with various driving maneuvers and results are evaluated

For many radar applications, a grouping of radar reflections that belong to the same object is needed. Unsupervised clustering algorithms are commonly used for this task. However,the number and density of measured reflections of an object depends on various parameters and therefore unsupervised algorithms often fail to identify all points that should be part of the same cluster. We propose a method to incorporate learned knowledge about the data into the clustering algorithm and show that this new method outperforms unsupervised approaches.

Semantic knowledge of the environment is one of the building blocks for autonomous driving. Therefore, it is desirable to maximize each sensor's semantic capabilities, since redundancies are often required by functional safety. Similar to camera sensors the radar sensor domain should move from more or less pure radar based measuring and tracking towards a semantic understanding of the surrounding. The talk shows our vision of a semantic instance segmentation of 360° spatio-temporal radar point clouds by means of deep learning. Our experiences after labeling several million radar samples are shared and we demonstrate how machine learning techniques can be combined with conventional signal and image processing algorithms.

The highly increasing road transport requires intelligent traffic solutions. The cost-effective passive RFID technology can help controlling the traffic flow in urban areas, can generate faster access into city centres, parking buildings and shopping malls and this way reducing the travelling time and emissions. In this talk the trends and applications for auto vehicle identification with passive RFID based on the newest generation of wirelessly connected readers, high gain selective antennas and secure trusted transponders will be presented.

With the advent of powerful algorithms that produce situational awareness from heterogeneous sensors, mobility is becoming 'intelligent', i.e. becoming aware of the situation, of what going on around a mobile platform. This rapid development occurs in all dimensions of mobility - on the ground, in the air, at sea, under water. This general trend opens up a vast variety of innovative use cases for expanding our knowledge in research projects, for protecting our way of life, for realizing new business opportunities. In using mobility more intelligently, sensor-generated situational awareness not only assists users in controlling mobile platforms, it enables platform automation up to autonomy, and is a crucial element in manned-unmanned teaming or swarming. Sensor Fusion Engines transform 'just in time' vast streams of data collected from a variety of sensors and non-sensor context data into pieces of information from which comprehensive situation pictures are produced. This goal is achieved by an optimized use of all available sensing, communications, and plat- form-related resources. The overall technological trends already push Sensor Fusion Engines into the position of a key enabling technology in many fields of application. In addition, Sensor Fusion Engines will become increasingly important by even more powerful processing units, by broadband and adaptive communication links, by high-precision and robust navigation systems, and by mobile platform technologies in general, on the one hand; and by progress in developing sophisticated mathematical algorithms on the other, often referred to as "artificial intelligence". We will motivate that progress in Fusion technology will evolve along the following lines: (1) The system design of mobile platforms will increasingly be dominated by artificially 'intelligent' data exploitation soft- ware in comparison to classical hardware components. (2) Individual sensors will be embedded into overall multiple sensor systems of mutually complementary and heterogeneous sensors. (3) Multi-functionality in sensor systems design is and will be a predominant factor, i.e. the shared use of the same sensing hardware to achieve several specialized goals. (4) Emphasis is to be placed on data integrity aspects of Fusion Engines. Besides Electronic Warfare issues, pressing in civil application as well, this issue comprises navigation and cyber security. (5) Emerging Sensor Fusion Engines will be inherently "cognitive" with respect to scenario and mission requirements and will massively exploit external knowledge bases. (6) Situational awareness provided by Fusion Engines is not only the key to reaching the goals of mobile activity more efficiently, but to reaching them also in an ethically acceptable and responsible way. We will address generic design principles of Sensor Fusion Engines, where the flow of data and information between human decision makers and the mission-related entities basically has two directions. First, there is a process harvesting from sensor data increasingly informative insight. Secondly, sensor resources management actively controls the information acquisition processes. This twofold flow of information, assessment and management, is driven by learning and reasoning algorithms. This more general picture is illustrated by a series of examples from ongoing research projects with unmanned ground, air, and underwater vehicles. They in particular show that advanced sensor technologies will (1) have to be seamlessly embedded into overarching Sensor Fusion Engines and Sensor Resources Management, (2) will have to widen the scope in that a combination of low-cost sensors to be fused may outperform a more expensive high performance sensor, (3) have to profit from advanced cognitive communications links in providing sensor data at different levels of preprocessing, (4) have to be cyber-safe and classically EW resistant, (5) have to be modeled realistically in terms of the accuracy and reliability of the sensor data they are delivering.

Automotive radars operate mainly in 77 GHz and 79 GHz frequency bands as 24 GHz radars are phasing out. These radars are integrated into bumpers of the car or mounted behind so called design emblems. These plastic materials need to be RF transparent to let the signals from the radar sensors pass without distortion. However, due to different paintings and coatings of bumpers low RF transparency and non-homogeneity can cause radars to malfunction and lower the radar performance. Testing the RF performance of materials requires typically a network analyzer and takes a long time. This paper presents a new measurement method to test and verify radome materials within seconds.

As the number of radio systems in modern vehicles increases, electromagnetic interferences between them become an issue. In-vestigating and testing new wireless devices regarding the im-munity against radio disturbances and the emission of unwanted signals is essential to reach high reliability in later applications. Usually, these tests are carried out in fully anechoic chambers, to achieve defined, undisturbed wave propagation mimicking free-space conditions. Because the degree of achieving nearly ideal conditions can strongly influence measurement results, interna-tional standards provide procedures to validate test facilities. This paper discusses the influence of an electrically large antenna measurement arch on the norm compliance of the Virtual Road Simulation and Test Area (VISTA) of the Thuringian Center of Innovation in Mobility at the Technische Universität Ilmenau, with focus on the frequencies of modern wireless communication systems. Time domain information of a received test signal is used to identify the reflections and their sources in the chamber. This approach provides access to the site voltage standing wave ratio (sVSWR) demanded by standards and presents a novel general-ized technique for reliable site qualification in consideration of installed structures.

In a simulation for stimulation of automotive radar an analytic connection is established between SBR-Phong and radar cross section (RCS) computations. Phong is adapted and normed for this purpose. RCS values are computed for a primitive- (sphere) and a realistic geometry (car) using the adapted Phong approach, and are compared against computations from a commercial field simulation software.

The number of antennas in cars is increasing by the day. The interaction of an automotive antenna with the car chassis presents a prominent design challenge. While patch antennas are resilient to the underlying chassis metal, they suffer from poor impedance matching bandwidth for a low-profile design. Dipole antennas provide higher bandwidths, but the radiation performance degrades significantly if the dipole is mounted less than a quarter-wavelength apart from the car chassis metal. This paper shows through consistent simulation and measurement results the significant benefits of using a dipole over a high impedance surface (HIS) when low-profile antenna designs for confined mounting locations close to the chassis metal are required. It was observed that the HIS-dipole delivered 35% total efficiency for a mounting location that was only 3 mm apart from the chassis metal. At the same time, a well-defined resonance with a sufficient bandwidth and a positive peak realized gain (2.6 dBi) for the LTE-2600 frequency band (2500...2700 MHz) were obtained. These results indicate a stable and reliable performance of plastic embedded HIS-dipoles in a car, and open new avenues for novel close-to-chassis automotive antenna integration locations.

The antenna is an important part of the automotive radar system and influences the sensor's performance, especially its angular behaviour. One possibility is the use of a conformal antenna array to reach a larger view angle. However, the Rayleigh criterion cannot give an appropriate measure for the resolution because the array elements have different view angles. In this paper the incoherent superposition of two targets is assumed according to the Rayleigh criterion. In the first step the separability of two targets is checked with the patterns of the antenna array for all target angle combinations. Coming from this, an assumption allows the use of the ambiguity function to estimate the resolution. This works well for the main beam and gives the advantage of a fast calculation.

We propose a design strategy for optimizing antenna positions in linear arrays for far-field Direction of Arrival (DoA) estimation of narrow-band sources in collocated MIMO radar. Our methodology allows to consider any spatial constraints and number of antennas, using as optimization function the Weiss- Weinstein bound formulated for an observation model with random target phase and known SNR, over a pre-determined Field-of-View (FoV). Optimized arrays are calculated for the typical case of a 77GHz MIMO radar of 3Tx and 4Rx channels. Simulations demonstrate a performance improvement of the proposed arrays compared to the corresponding uniform and minimum redundancy arrays, even if the latter have a larger aperture.

Radar sensor has been an integral part of safety critical applications in automotive industry owing to its weather and lighting independence. The advances in radar hardware technology has made it possible to reliably detect objects using radar. The highly accurate radar sensors give multiple radar detections per object. In this work, a postprocessing architecture is presented which is used to cluster and track multiple detections from one object in practical multiple object scenarios. Furthermore, the framework is tested and validated with various driving maneuvers and results are evaluated

For many radar applications, a grouping of radar reflections that belong to the same object is needed. Unsupervised clustering algorithms are commonly used for this task. However,the number and density of measured reflections of an object depends on various parameters and therefore unsupervised algorithms often fail to identify all points that should be part of the same cluster. We propose a method to incorporate learned knowledge about the data into the clustering algorithm and show that this new method outperforms unsupervised approaches.

Radar has proven to be an advantageous sensor principle in many traffic related applications. In infra- structures, it has fulfilled its promise to be independent from of environmental conditions to a remarkable degree while at the same time being capable of covering a large field of view per sensor. This talk covers how radar sensors are used today and what particular properties make them useful in today's applications. It shows both advantages and future challenges of radar sensing from an infrastructure perspective. Furthermore, we present examples of technological aspects with regards to novel use cases like generation of comprehensive occupancy maps.

The measurement of the filling level of liquid and solid goods in storage, process, and other tanks and silos is an important task in many industries. Examples of liquids are oil, gasoline, various chemicals and pharmaceuticals, water, and beverages. Also, a large variety of bulk solids can be given, for example fine granulated solids like flour, sugar, sand, grains, and various powders. Examples of more rough solids are pellets, coal, grit, stones, and others. Resulting from the different electrical properties (permittivity and conductivity) and the different sizes and structures, the reflection, scattering, and attenuation properties are strongly different. Radar sensors operating at frequencies up to 24 GHz are already well-established for level measurement applications. Radio regulations and the latest state-of-the-art technology have now also made radar level meters at high frequencies in the 80 GHz range commercially available. In this contribution, first the measurement scenarios and conditions given with liquids and bulk solids are both discussed and compared to each other. Secondly, criteria and technical backgrounds for a suitable choice between 24 GHz and 80 GHz radar systems dependent on the specific level measurement application are motivated and given. Thereafter, the realization of components for according radar sensors is presented and discussed, with a focus on the radio-frequency (RF) front-end electronics using monolithic microwave integrated circuits (MMIC) and different radar antenna concepts. For the latter, the given harsh conditions in the application are a challenge, as will be discussed. Finally, some exemplary results obtained from practical measurements with the discussed radar systems are presented and analyzed.

This paper investigates the potential of applying synthetic aperture radar (SAR) to automotive applications using a short range vehicle-mounted Frequency Modulation Continuous Wave (FMCW) radar. The Range Migration Algorithm (RMA) is adopted in this work for SAR image formation as it is ideal for focusing range curvatures of targets at short ranges. The first part of this paper gives a brief overview of RMA based strip map SAR processing of FMCW radar signals. In the latter part, collected data of real world stationary targets such as cars, poles and buildings are processed providing high resolution SAR images of different synthetic aperture lengths. The obtained images are presented, discussed and compared in terms of cross range accuracy and resolution to show that high resolution target imaging can be achieved using automotive SAR.

In this contribution results from radar based Doppler measurements for rain drop detection will be presented and discussed. It will be pointed out that using such sensors not only the velocity but also the incident angle of the impinging rain drop can be determined which is an additional relevant parameter for an environmental sensor. Furthermore, the importance of simulation will be addressed for an enhanced understanding of such multi-particle problems.

Future high-resolution radars enable new functionalities in advanced driver assistance systems, relying on fast and reliable extraction of properties of vehicles on the road. A critical property for the prediction of trajectories and the assessment of potentially dangerous situations is that of the actual motion - the velocity vector and yaw rate - of observed objects. In this paper, an approach to distinguish linear from non-linear motions as well as estimating the yaw rate using only a single radar sensor is presented and evaluated via measurements.

In this paper, we consider a type of MIMO-SAR architecture in which the elements are arranged along the axis of platform motion. We show that this architecture can be used to improve the robustness and accuracy of angular estimation compared to single-element SAR given array location uncertainty. We compare three different architectures for side-looking radars (SAR, SIMO, and SIMO-SAR) and show that in the presence of array location uncertainty, SIMO-SAR proves more resilient. Measurement results using a commercial automotive radar module have been performed to verify this effect.

High resolution radar imaging is highly desired for autonomous vehicle systems but the complexity and cost have been prohibitively expensive. Coded Aperture Radar (CAR) is a technique that simplifies the RF hardware, utilizing single-bit phase shifters to obtain angular information, and enables digital beamforming with only a single transceiver. This paper describes the development and measured results of a CAR sensor developed for 77 GHz operation that utilizes 128 array elements to digitally synthesize directional beams. The authors quantify the sensor performance in cluttered environments.

This paper provides an approach for the evaluation of coherent interference potentials and sensitivities of fast chirp radars which focuses on the influence of ramp patterns and timing effects. The proposed approach combines a simulation study with synchronized radar measurements in order to determinate the ghost target detection rate in specific scenarios. Furthermore the developed method is applied to compare the performance of particular time hopping algorithms for fast chirp radars regarding their interference mitigation efficiency.

An automotive radar sensor for cocoon functions or automated parking requires very small dimensions to access new mounting positions like B-pillar and side skirts. To minimize the dimensions of radar sensors, new concepts are necessary. A new system approach for radar sensors is presented. The new radar sensor system is divided in two major units. The sensor unit consists of a small serializer board and an LTCC (Low Temperature Cofired Ceramic) miniature frontend. The external radar ECU (Electrical Control Unit) provides the signal processing performance and the power supply for the sensor unit. For the automotive radar band (76-81 GHz), RX- and TX antennas have been simulated, manufactured and the radiation pattern has been measured.