Conventional radar segmentation research has typically focused on learning category labels for different moving objects. Although fundamental differences between radar and optical sensors lead to differences in the reliability of predicting accurate and consistent category labels, a review of common radar perception tasks in automotive applications reveals that determining whether an object is moving or static is a prerequisite for most tasks. To fill this gap, this study proposes a neural network (NN)-based solution that can simultaneously segment static and moving objects from radar point clouds. Furthermore, since the measured radial velocity of static objects is correlated with the motion of the radar, this approach can also estimate the instantaneous 2-D velocity of the moving platform/vehicle (ego-motion). Notably, despite performing dual tasks, the proposed method employs very simple yet effective building blocks for feature extraction: multilayer perceptrons (MLPs) and recurrent NNs (RNNs). In addition to being the first of its kind in the literature, the proposed method also demonstrates the feasibility of extracting the information required for the dual tasks directly from unprocessed point clouds, without the need for cloud aggregation, Doppler compensation, motion compensation, or any other intermediate signal processing steps. To measure its performance, this study introduces a set of novel evaluation metrics and tests the proposed method using a challenging real-world radar dataset, RadarScenes. The results show that the proposed method not only performs well on the dual tasks but also has broad application potential in other radar perception tasks. More qualitative results can be viewed here: https://youtu.be/3ejS1chSvQ8?si=uGRugVA63BCyvNBV

To characterize atmospheric turbulence, the Doppler moments are estimated by weather radars. However, moment accuracy is highly sensitive to radar transmission parameters such as pulse repetition time (T_s) and number of pulses (N_p), which affect Doppler ambiguity and estimation variance. Traditional fixed-parameter radars face trade-offs between aliasing and measurement precision. This paper proposes an adaptive radar framework that dynamically adjusts T_s and N_p on a per-scan basis to improve Doppler moment estimation at a single resolution cell level. Inspired by the Fully Adaptive Radar (FAR) concept, the method also includes a novel multi-lag Doppler width estimation scheme. Results demonstrate enhanced estimation accuracy, enabling better responsiveness to localized and non-stationary weather conditions.

The impact of cross-polar coupling and a finite available number of echo samples on the data quality of polarimetric weather radar data is investigated. A system model and a two-step simulation approach are proposed to simulate realistic time series of the received weather radar signals. The proposed simulation methodology is applied to a rainfall scenario to illustrate the data quality of two widely used weather radar measurement schemes. Through simulation of the bounds within which 90% of the resulting estimates fall, it is demonstrated that the finite sample size introduces deviations that limit the data quality, even in designs with low cross-polar coupling.

A data driven method is proposed to obtain free space segmentation using automotive radar point clouds. It aggregates automotive radar detection points from multiple timestamps, projects them into a Birds-Eye-View grid-based representation, and applies a semantic segmentation Neural Network (NN) to classify each grid for free space segmentation. A lidar based supervision is used to generate the ground truth for training. Moreover, debris objects are manually annotated to enable the NN model to learn to detect these uncommon objects. Experimental results on a proprietary 4D Imaging Radar dataset demonstrate that the proposed method gives improved free space segmentation as compared to other baseline methods.

In this contribution, we analyze machine learning-assisted solutions to tackle real-time fault detection in large-scale active phased array antennas. The challenge of integrating the circuit component nonlinearities and mutual coupling effects in fault-finding methodologies is addressed. A novel machine learning (ML) solution based on an array theory-enhanced neural network (NN) is proposed. To address the practical constraints of large array measurements, sparse far-field (FF) measurements are considered. The method is experimentally verified by applying it to fault detection in a 64-element planar phased array prototype operating at 26 GHz with far-field measurements collected by a fixed high-speed multi-probe pattern acquisition setup, the Antenna Dome. Significant improvement in fault prediction performance over a conventional Genetic Algorithm (GA) based heuristic approach with improvements of up to 40%, 25%, and 20% in predicting 8-, 4-, and 2-element faults, respectively, is demonstrated. The robustness of the proposed performance with respect to the number of 3D spatial field sampling points is shown, offering efficient diagnostics with in-field pattern sampling compatibility and low hardware complexity.

The feasibility of using hybrid additive manufacturing (AM) to produce low-cost, phased array antennas for SatCom applications is investigated in this work. A 4x2 planar array comprised of dielectric resonator antennas (DRAs) operating in the fundamental mode within the K-band SatCom downlink band (17.7-21.2 GHz) has been designed, fabricated and measured. The impact of print settings on the material properties is assessed and incorporated into the antenna design process. Manufactured prototypes of the single element and the planar array geometry are experimentally verified in terms of their scattering parame-ters and far-field radiation patterns. The potential of combining hybrid AM and DRA technologies for future mmWave phased array developments is highlighted by the presented results.

The effect of amplifier-related signal amplitude compression in orthogonal time-frequency space (OTFS) waveform for radar and communications systems is considered. A novel approach to OTFS waveform generation is proposed, where complementary sequences are used with the Zak transform to encode delay-Doppler symbols and form an OTFS time-domain signal with a constant envelope. The high peak-to-average power ratio (PAPR) of conventional OTFS can cause amplifier saturation, leading to spectral noise and performance degradation in both communication and radar systems due to amplitude clipping. This issue can be critical in dual-function radar and communication applications, where high power may be crucial in both use cases. The proposed waveform, namely, constant modulus OTFS (CM-OTFS), offers an alternative to standard OTFS when high-power or low-cost amplification is required. The sensing and communications performances of CM-OTFS are evaluated through numerical simulations and compared with pristine and amplifier-distorted OTFS waveforms. CM-OTFS demonstrates slightly degraded sensing performance and lower communication rate than pristine OTFS but outperforms amplifier-distorted OTFS signals. The performance of CM-OTFS is evaluated through radar and communication simulations, as well as radar measurements using the waveform-agile PARSAX radar.

The problem of estimating breathing rates and detecting apnea events with radars located at an elevated and tilted position is considered in this paper. This is particularly relevant in psychiatric clinics, where radars (or other sensors) must be installed out of reach of patients. In this work, a feasibility study is presented, using an experimental setup with two 60 GHz Frequency-Modulated Continuous Wave (FMCW) radars placed at 2.7 m height at a tilted angle towards the participants, and one radar at 1 m height looking straight to the participants, who are sitting and lying on the floor. A new comprehensive dataset with 30 participants and 7 activities was collected with this setup. Using phase extraction and filtering, the work presents an apnea detection probability of up to 90 % with an elevated radar, and comparable mean error rates for breathing estimation of below 2 respirations per minute (rpm) for all radars. The results show that the respiration data and apnea detection from all radar positions are comparable. This proves the feasibility of the proposed radar deployment positions, benefiting application fields such as psychiatric care.

A novel concept to mitigate the unintended transmission and reception of signals at the third harmonic of dielectric rod antennas is proposed. To suppress the third-harmonic radiation, an additive-manufactured photonic bandgap material is used in the dielectric rod antenna. The antenna has been designed to operate at the frequency of 5 GHz and exhibit a bandgap at the third-harmonic frequency of 15 GHz. The design of the material is explained via the band diagram of the bandgap material and imperfections introduced due to the additive manufacturing (AM) process considered. Numerical simulations of dielectric rod antenna prototypes fed via a rectangular waveguide (RWG) with and without harmonic suppression are carried out to confirm the operational principle. The design proposed is verified experimentally. The manufactured antenna is characterized in terms of its input reflection coefficient and far-field radiation properties. The obtained experimental results agree well with predictions obtained through simulations and confirm the third-harmonic suppression capability of the antenna.

An antenna array with dual-functionality - electromagnetic radiation and thermal cooling - is proposed. An iterative array design procedure is developed to improve cooling, mutual coupling, side lobe levels, and gain levels in dual-functional antenna arrays with adaptive beam steering. Heatsink-attached patch elements are combined with complementary split ring resonator (CSRR) structures in between the elements, resulting in a novel modular heatsink antenna array. Based on the proposed design, the beam scanning performances of four-element and eight-element linear arrays at 26 GHz are studied. A conventional shorted patch antenna array is used for benchmarking. Through thermal and electromagnetic simulations, it is demonstrated that the proposed antenna array decreases the maximal array temperature by more than 40°C as compared to the benchmark. Moreover, the new design resolves the pattern performance degradation problems in heatsink arrays, while approaching to the electromagnetic performance of the benchmarked array.

An overview of polarimetric sensing and its growing application in automotive radar systems is presented. While polarimetric techniques are extensively used in fields like weather monitoring and target imaging, their integration into automotive radar presents unique challenges, particularly in calibration and measurement accuracy across wide scanning angles. This paper reviews key polarimetric principles and their use in different applications, with a focus on current automotive radar implementations, and the calibration challenges posed by off-broadside measurements. Future research directions for improving polarimetric accuracy in dynamic automotive environments are also discussed.

A novel electro-thermal model at the unit-cell level for active phased arrays is proposed to establish a link between the power amplifier (PA) output signal and its junction temperature. The iterative model is developed in four stages: (i) calculation of the PA output, (ii) computation of the dissipated power of the PA, (iii) thermal simulation for temperature assessment, and (iv) update on the PA output with the new temperature. These steps are repeated until the PA temperature is convergent. The use of the model is demonstrated with an amplitude-tapered 16-element array of single-stage class-A amplifier and patch antenna unit-cells at 2.14 GHz. It is observed that the inclusion of the temperature effects in the array pattern causes around a 2 dB drop in the radiated power, while the peak side lobe level (SLL) increases by up to 7.75 dB for 40 dB Taylor tapering.

The problem of enhancing angular resolution capability using a non-coherent radar network has been considered. A 2D-Localization approach based on block-FOCUSS is proposed to overcome the near-field effects. In this method, range-angle coupling due to the large baseline between radar units is taken into account. The performance of the proposed method is compared with two state-of-the-art direction-of-arrival estimation methods which are based on block-FOCUSS and block-OMP.

The problem of enabling adaptive capabilities in the context of weather radar is considered in this paper. Inspired by the cognitive radar framework, an approach based on Reinforcement Learning (RL) is formulated to deal with the monitoring of multiple storm cells moving near a potential area of interest. The approach aims to dynamically adjust the radar waveform bandwidth, and consequently maximum measurable range and range resolution, in order to provide the best monitoring based on a purposely-defined reward function. The approach is successfully validated with a simulator developed in Python & StoneSoup. Results demonstrate that the proposed method outperforms traditional fixed-scan ('sit and spin') strategies commonly used in weather radar operations.

The problem of radar-based multi-target tracking for indoor human monitoring is considered. Tracking and counting the number of people moving as a group is particularly challenging as multiple individuals are close together and their radar signatures are mixed. A transformer-based classification approach for counting the number of grouped people is proposed. The Neural Network model is trained with selected features from the spatial domain and Doppler frequency domain, which are concatenated over multiple frames to form a sequence for the transformer network. Compared to statistical classifiers, the self-attention mechanism allows transformers to capture feature long-term dependencies. The proposed classifier is integrated into a tracking pipeline in order to monitor the position and number of people in the grouped targets. The method proposed is experimentally verified using a 24 GHz Frequency Modulated Continuous Wave (FMCW) radar with 250MHz bandwidth. Despite the relatively coarse range resolution, the proposed method achieves 92.5% accuracy in these initial tests. Furthermore, the method performances and related accuracy is analyzed according to various parameters.

The reconstruction of Embedded Element Patterns (EEP) in an array of antennas is presented as an attempt to understand the effects of Mutual Coupling (MC) among the antenna elements. The EEP far fields are modeled with a weighted sum of orthonormal spherical harmonic basis functions (explained by their mode numbers). The weight of each significantly contributing mode is estimated by the Singular Value Decomposition (SVD) approach. A practical example of a non-uniform array of circular patch antennas is considered for analyzing the significant contributing modes. Although the values of the weights of the modes are different for each antenna element in the array, the contributing modes remain very similar for all the elements, indicating that the mutual coupling can be sufficiently explained by a limited number of variables (weights) corresponding to limited spherical harmonic modes. The reconstructed patterns are validated by comparing them with full-wave simulation results.

A novel over-the-air (OTA) measurement setup for simultaneous and near-real-time characterization and mapping of the phased array radiation patterns to the error-vector-magnitude (EVM) is introduced. A cost-effective software-defined radio (SDR) system, including a Pluto SDR for both transmission and reception, is employed. For setup demonstration, the performance of a 4 × 4 active phased array antenna (APAA) is studied at 26.5 GHz. Calibration of the transmitter results in an EVM of 0.5%, proving the accuracy of the measurement system. The over-the-air measured EVM of the APAA is 1.2%. The proposed setup and approach provide deep insights into how array radiation characteristics affect signal quality, advancing the evaluation and development of APAAs for next-generation wireless communication systems.

The problem of radar-based, continuous Human Activity Recognition (HAR) has been studied in this work. A fixed-window segmentation method based on dual timescales has been proposed to tackle this challenge. The method is experimentally validated on a challenging publicly available dataset with 14 participants and 9 activities, and is compared to reference works from the literature. L1PO validation of the method yields a test accuracy and macro F1-score of 87.5 % and 80.1 % respectively.

The problem of diminished unambiguous target velocity interval induced by the time-division-multiplex mode (TDM) of multiple-input-multiple-output (MIMO) frequency-modulated continuous-wave (FMCW) automotive radar has been explored. A novel MIMO antenna array activation mode and a parametric approach for Doppler de-aliasing based on a two-step cross-entropy optimization are proposed. The TDM Doppler signal model has been derived, and a novel two-step cost function is proposed to achieve robust and efficient estimation. In contrast to the state-of-the-art method, the method proposed does not need multiple overlapped antennas and can resolve multiple targets in the same range and Doppler bins. The proposed method has been verified with numerical simulations with different parameter settings, as well as experimental data from a radar target simulator.

In this article, the classification of dynamic vulnerable road users (VRUs) using polarimetric automotive radar is considered. To this end, a signal processing pipeline for polarimetric automotive MIMO radar is proposed, including a method to enhance angular resolution by combining data from all polarimetric channels. The proposed signal processing pipeline is applied to measurement data of three different types of VRUs and a car, collected with a custom automotive polarimetric radar, developed in collaboration with Huber+Suhner AG. Several polarimetric features are estimated from the range-velocity signatures of the measured targets and are subsequently analyzed. A Bayesian classifier and a convolutional neural network (CNN) using these estimated polarimetric features are proposed and their performance is compared against their single-polarized counterparts. It is found that for the Bayesian classifier, a significant increase in classification performance is achieved, compared to the same classifier using single polarized information. For the CNN-based classifier, utilizing the distribution of polarimetric features of the target’s range-velocity signatures also increases classification performance, compared to its single-polarized version. This shows that polarimetric information is valuable for classification of VRUs and objects of interest in automotive radar.