Estimation problems in wireless sensor networks (WSNs) typically involve collecting and processing data from distributed sensors at the fusion center to infer the state of an environment. However, not all measurements contribute equally to estimation accuracy. In this work, we incorporate the concept of ordered transmission into sequential estimation to select the most informative measurements from different sensors, while ensuring the desired estimation quality. We analyze a general estimation problem with different estimator choices and derive stopping rules for collecting measurements. Then, we derive the expected number of transmissions required for our ordered transmission-based sequential estimation scheme and compare it with that of a conventional sequential estimation scheme with unordered transmissions. To validate the proposed protocol, we apply it to a radar-based WSN for target localization and velocity estimation, designing an ordered transmission strategy and a sequential stopping rule. Simulation results show that our protocol requires fewer transmissions compared to conventional sequential estimation while maintaining similar estimation accuracy in general WSNs. In a radar-based WSN, the proposed protocol achieves reliable estimation with reduced communication overhead and improved response time.

Compressive sensing (CS) is key to reduce the overhead in estimating sparse high dimensional channels at millimeter wave or terahertz frequencies. The channel measurements in CS are usually perturbed by random phase errors, commonly modeled as a Wiener process, at the oscillators. CS algorithms that ignore such phase errors fail to accurately estimate the channel. In practice, the phase errors are similar within a batch of measurements acquired in a short burst and the errors vary significantly across different batches, resulting in partially coherent measurements. We develop a message passing-based channel estimation algorithm that exploits the sparse structure of the channel together with the Wiener statistics of the phase errors. To this end, we absorb the phase errors into the sparse channel, and introduce three hidden variables to model its support, magnitude, and phase. We derive the message flows between these variables while incorporating Wiener phase noise statistics. Finally, we use alternating optimization to decouple the sparse channel and the phase errors from the vector estimated with our message-passing technique. Using simulations, we show that the proposed algorithm achieves better channel reconstruction than comparable benchmarks.

Direction-of-arrival (DoA) estimation is a key operation in 5G radios, radars, and sonars. While large receive arrays enable high-resolution DoA estimates, their fully digital implementation consumes significant power. This paper demonstrates DoA estimation with a switched receive array, which consumes less power than its fully digital counterpart. The DoA is estimated in two stages. First, the sector in which the source lies is estimated by mechanically steering a wide beam. Then, the DoA within the identified sector is estimated electronically using our switched receive array. We formulate an integer program to optimize the configuration of switches at the receiver, resulting in low-aliasing artifacts within the identified sector. Using our custom 40 KHz ultrasound receive array, we demonstrate DoA estimation with our optimized switch configuration. Experiments and numerical results show that the error in the estimated DoA is smaller with our optimized switch configuration than with a randomly chosen configuration.

Millimeter wave (mmWave) systems, currently employed in 5G and IEEE 802.11ad/ay devices, enable high data rates through wide bandwidths and directional communication. However, high carrier frequencies used in these systems result in a higher phase noise than lower frequency systems. This paper investigates the problem of spatial channel estimation in the presence of severe phase noise, which manifests as partially coherent phase perturbations in the observed channel measurements. In this model, phase noise remains relatively constant within a packet but varies substantially across packets. Under such partially coherent phase noise, we first develop two computationally efficient on-grid algorithms to estimate narrowband mmWave channels: Partially Coherent Matching Pursuit (PCMP) and Enhanced Partially Coherent Matching Pursuit (EPCMP), assuming a known channel sparsity. Both algorithms exploit the sparse structure in mmWave channels, enabling a significant reduction in training overhead while achieving good estimation performance. The main difference between PCMP and EPCMP is how the sparse channel support is identified. The EPCMP algorithm can achieve better estimation performance at the cost of increased computational complexity compared to the PCMP algorithm. We then relax the known-sparsity assumption, adapt the proposed algorithms accordingly, and further extend them to the wideband case for an unknown sparsity. Additionally, we derive sufficient conditions to recover a support element with proposed algorithms. Simulation results demonstrate the advantages of our methods over comparable channel estimation benchmarks.

Occupancy grid mapping is a common approach to support automotive driving perception capabilities. We present an occupancy grid estimation algorithm using sensor point-cloud measurements aided by side information from other sensing modalities like cameras. This prior side information is in the form of an erroneous occupancy map estimate, referred to as prior support information. Specifically, we extract a prior map using you only look once (YOLO) object detection on camera images. A sparse Bayesian learning-based mapping algorithm is designed with a modified hierarchical model to incorporate this prior. Experiments done on public real-world driving datasets, nuScenes and RADIATE, demonstrate that our approach achieves better target detection and scatter noise reduction than the state-of-the-art methods. Furthermore, our method seamlessly works on the two datasets although we train YOLO only using camera images from nuScenes.

We consider the problem of generating automotive radar super-resolution maps from low-resolution radar maps and camera images. This problem is relevant in automotive driving for synthetic sensor data generation to support improved environmental perception. We propose a radar super-resolution sensing approach based on multimodal data fusion between low-resolution radar range-azimuth (RA) maps and aligned camera images. Our method employs a U-Net-based autoencoder architecture enhanced with visual features extracted from a pre-trained ResNet50 encoder, enabling the model to generate high-resolution RA maps that approximate ground truth radar data. We evaluate the proposed method on the RADIal and RaDICaL datasets, which cover diverse driving environments and radar configurations. Quantitative and qualitative results demonstrate that our approach outperforms a baseline model and prior state-of-the-art methods, particularly in resolving fine spatial details in scenarios with closely-spaced vehicles and pedestrians.

In-phase and quadrature-phase (IQ) imbalance in high-frequency systems distorts received measurements, causing standard channel estimation algorithms that ignore this mismatch to fail. In this paper, we develop an augmented compressed sensing (CS) model to jointly estimate the sparse channel and the IQ imbalance (IQI) parameter in the form of an augmented CS vector. This vector exhibits group sparsity, which is exploited using our tailored paired-support orthogonal matching pursuit (PSOMP) algorithm. Finally, the estimate from PSOMP is decomposed to determine the channel and the IQI parameter. Numerical results show that our method achieves better support recovery and a lower error in the estimated channel than the baselines.

Digital radars with low-resolution analog-to-digital converters (ADCs) have attracted attention as a solution to reducing the high digital processing complexity and power consumption at the receiver. Radars employing low-resolution ADCs, however, have a limited dynamic range, due to which high-radar cross section (RCS) targets mask low-RCS targets. The masking occurs because the quantized output is primarily determined by returns from high-RCS targets. To enhance the dynamic range of such radars, we propose to operate the ADC at a high resolution in the initial slow-time slot of each radar frame. The resulting high-resolution measurements are used together with the known Doppler statistics of dominant targets to construct a dither signal, which is used as a quantization threshold to acquire low-resolution ADC measurements in the subsequent slow-time slots. By incorporating situation awareness in the form of Doppler statistics, our dither signal can suppress returns from strong targets, effectively unmasking weak targets with low-resolution measurements. We analyze system performance in terms of the probability of detection and show that the proposed approach outperforms existing methods in enhancing the detection of weak targets. The simulations demonstrate that our method significantly improves target detection and reduces the normalized mean square error (NMSE) in the estimated radar channel over comparable benchmarks.

Block compressed sensing (BCS) alleviates the high storage and memory complexity with standard CS by dividing the sparse recovery problem into sub-problems. This paper presents a Welch bound-based guarantee on the reconstruction error with BCS, revealing that sparse recovery deteriorates with more partitions. To address this performance loss, we propose a data-driven BCS technique that leverages correlation across signal partitions. Our method surpasses classical BCS in moderate SNR regimes, with a modest increase in storage and computational complexities.

Digital radars with low-resolution analog-to-digital converters (ADCs) can reduce digital processing complexity and power consumption but suffer from limited dynamic range. The poor dynamic range causes high radar cross-section (RCS) targets to mask low-RCS ones. To mitigate this issue, we propose operating the ADC at a high resolution during the initial slowtime slot of each radar frame. The high-resolution measurements are used to estimate the range and RCS of dominant targets, which, along with their known Doppler statistics, are used to construct a dither signal. This dither signal is then employed to acquire low-resolution ADC measurements in the subsequent slow-time slots. With the proposed receiver architecture, our method suppresses strong target returns in the low-resolution measurements, effectively unmasking weak targets. Simulations demonstrate significant improvements in target detection and reduced normalized mean square error in radar channel estimation compared to existing benchmarks.

Light detection and ranging (LiDAR) is used in robots and in automotives to obtain the perception of the surrounding environment. Traditional spinning LiDARs scan the environment uniformly along all angular directions by operating at a constant rotational speed, with fixed sensing parameters throughout a rotation. Such a sensing approach, however, is suboptimal when information about static obstacles in the environment is available at the LiDAR. In this work, we introduce ELLAS, a first-of-its-kind spinning LiDAR system that dynamically adapts its range and resolution over the field of view. This adaptation is achieved by optimizing the ranging parameters at the LiDAR and the instantaneous rotational speed of the spinning platform to the location of static objects in scene topology maps. With the optimized settings, ELLAS results in a longer range along directions where static obstacles are farther away and achieves a higher angular resolution around those directions.

Radar is a key technology in automotive driving for target detection and perception. In this work, we leverage prior environmental information in the form of occupancy maps to design space-time codes for a fully digital MIMO radar. We transform this design problem into the optimization of spatial beamforming gains and time-domain codes. The beamforming gains are optimized to enhance the strength of returns from cells associated with a higher uncertainty of occupancy. The timedomain codes are optimized to minimize the correlation between returns of targets within the drivable space. We validate our method on the nuScenes dataset to show that the designed spacetime codes achieve higher detection rates than designs that do not rely on prior information from occupancy maps.

Phase jitter at oscillators in high-frequency wireless systems perturbs the phase of the acquired channel measurements. As a result, standard sparse channel estimation algorithms that ignore phase errors fail. In this paper, we consider a frame structure in which channel measurements are acquired over two packets. Our model assumes that the phase errors are nearly constant over a packet and they change considerably across the packets, leading to partially coherent channel measurements. In this paper, we develop a message-passing-based technique that leverages the partially coherent structure in the measurements for sparse channel estimation robust to unknown phase offset across the packets. Simulation results show that our approach achieves a lower mean-squared error in the reconstructed channel than comparable benchmarks.

Light detection and ranging (LiDAR) plays a crucial role in machine perception for advanced driver assistance systems. Existing LiDARs, however, do not adapt their sensing strategy to complement driver's perception. We demonstrate a novel LiDAR prototype that dynamically adapts its range and resolution over the field of view, according to real-time driver gaze. Our gaze-aware LiDAR emphasizes scanning peripheral zones the driver may overlook, i.e., critical areas during driving. Our demonstration showcases enhanced perception, highlighting the potential of hybrid human-machine sensing for safer driving.

Automotive LiDARs typically have a uniform scanning range over their field of view (FoV). Such a range profile does not account for the varying risk of misdetecting targets in different regions. For instance, prioritizing crosswalks in a LiDAR scan is crucial, as the financial consequences of missing a pedestrian far exceed that of overlooking a distant vehicle. In this paper, we construct a spatial risk map that quantifies the risk of misdetecting targets across different regions around the vehicle. Our risk map incorporates lane semantics, knowledge about previously identified objects, and their potential trajectories. We use this risk map to adapt the LiDAR's scanning range over different sectors in its FoV. Simulations on nuScenes episodes demonstrate that our misdetection risk-aware design reduces the effective risk by about 40% compared to a standard LiDAR.

Channel estimation can lead to a substantial training overhead in millimeter wave (mmWave) and terahertz (THz) systems employing large arrays. Prior work has leveraged channel sparsity at these frequencies to reduce this overhead. Most of the sparsity-aware algorithms, however, assume perfect phase coherence in the channel measurements, which is disrupted due to phase noise. Due to the errors induced by phase noise, standard sparse channel estimation algorithms assuming perfect phase coherence can fail. In this paper, we consider a frame structure in which the channel measurements are acquired over multiple packets. Our model assumes that the phase errors remain constant within a packet and vary considerably across different packets, leading to partially coherent channel measurements. We develop a message passing-based technique for sparse channel estimation under such partially coherent phase errors and show that our approach achieves a lower channel reconstruction error than comparable benchmarks.

We address the problem of estimating a binary occupancy grid map by fusing point cloud data from radar and LiDAR sensors for automotive driving perception. To achieve this, we introduce two measurement models for fusion and formulate occupancy mapping as sparse vector reconstruction from the set of radar and LiDAR measurements. The first model, called common sparse fusion, jointly estimates a common map from all sensor measurements. The second model, called common innovative sparse fusion, assumes a shared map and an innovation component (error collector) for each sensor modality’s measurements. This approach enhances the robustness of occupancy map estimation against potential sensor mismatch and calibration errors, and inconsistencies between the two modalities. We use the pattern-coupled sparse Bayesian learning (PCSBL) algorithm to recover maps, leveraging the inherent sparsity and spatial dependencies in automotive occupancy maps. Numerical experiments on the public RADIATE dataset show that our feature-level fusion models outperform single-modality mapping and decision-level fusion models in detecting drivable areas and targets. Furthermore, statistical results with corrupted LiDAR data establish that our common innovative sparse fusion model is robust against unreliable sensor data

Beam acquisition is key in enabling millimeter wave and terahertz radios to achieve their capacity. Due to the use of large antenna arrays in these systems, the common exhaustive beam scanning results in a substantial training overhead. Prior work has addressed this issue by developing compressive sensing (CS)-based methods which exploit channel sparsity for faster beam acquisition. Unfortunately, most CS techniques employ wide beams and suffer from a low signal-to-noise ratio (SNR) in the channel measurements. To solve this challenge, we develop an IEEE 802.11ad/ay compatible technique that takes an in-sector approach for CS. In our method, the angle domain channel is partitioned into several sectors, and the channel within the best sector is estimated and then used for beamforming. The essence of our framework lies in the construction of a low-resolution beam codebook to identify the best sector and in the design of a CS matrix optimized for in-sector channel estimation. Our beam codebook illuminates distinct non-overlapping sectors and can be realized with low-resolution phased arrays. We show that the proposed codebook results in a higher received SNR than the state-of-the-art sector sweep codebooks. Furthermore, our optimized CS matrix achieves a better in-sector channel reconstruction and a higher achievable rate than comparable benchmarks.

Millimeter-wave radar is a common sensor modality used in automotive driving for target detection and perception. These radars can benefit from side information on the environment being sensed, such as lane topologies or data from other sensors. Existing radars do not leverage this information to adapt waveforms or perform prior-aware inference. In this paper, we model the side information as an occupancy map and design transmit beamformers that are customized to the map. Our method maximizes the probability of detection in regions with a higher uncertainty on the presence of a target. Simulation results on the nuScenes dataset show that the designed beamformer achieves substantially higher detection rates than a conventional omnidirectional beamformer for the same transmitted power.

Occupancy grid maps provide information about obstacles and available free space in the environment and are crucial in automotive driving applications. An occupancy map is constructed using point cloud data from sensor modalities such as light detection and ranging (LiDAR) and radar used for automotive perception. In this article, we formulate the problem of estimating the occupancy grid map using sensor point cloud data as a sparse binary occupancy value reconstruction problem. Our proposed occupancy grid estimation method, based on pattern-coupled sparse Bayesian learning (PC-SBL), leverages the sparsity and spatial dependencies inherent in occupancy maps typically encountered in automotive scenarios. The proposed method shows enhanced detection capabilities compared to two benchmark methods based on performance evaluation with scenes from the nuScenes and RADIal public datasets.