This paper presents a 12-bit 10-GS/s time-interleaved (TI) ADC with distortion-cancelled track-and-hold amplifier (THA). A half-swing input buffer and ring amplifier (ringamp)-based output buffers are implemented to reduce THA nonlinearity and power consumption. Static and Dynamic distortion cancellation schemes are proposed to enhance THA linearity without imposing extra circuit loading on the buffers. These schemes leverage the opposite polarity of nonlinearities from each circuit component to realize cancellation—static distortion cancellation between the input buffer and output buffer, and dynamic distortion cancellation between the input buffer and sampling network. Fabricated in a 28-nm CMOS process, the prototype ADC consumes 198.1 mW, including 63.9 mW for the THA. At Nyquist input frequency, the ADC achieves 50.48-dB SNDR and 69.91-dB SFDR, equivalent to a Walden FoM of 72.6 fJ/conv.-step and a Schreier FoM of 154.5 dB, demonstrating outstanding linearity and power efficiency.

4D millimeter-wave radar has attracted significant research interest in 3D object detection for autonomous driving as its stability in adverse conditions. However, the sparsity and noise nature of the radar hinders their performance. The virtual-points-based methods with radar-camera fusion may alleviate these limitations. Nevertheless, these approaches primarily focus on the role of hybrid points (including virtual and raw points) in the radar branch and fail to exploit their full potential in the image branch. Hence, we introduce a novel 4D radar and camera fusion network, named HPAFusion for 3D object detection. Specifically, we first propose the Hybrid Points-assisted Depth Net (HPDN), which projects hybrid points onto the image coordinate system and uses voxel grids and points backbone encoding them to assist the image in obtaining depth information. Then, we present the Dual Attention Fusion Module (DAFM), which employs the uniform channel attention and spatial attention for facilitating cross-modal interaction and improving fused features. The proposed HPAFusion is validated on the TJ4DRadSet and View-of-Delft (VoD) datasets. Experimental results demonstrate that HPAFusion effectively fuses camera and radar with hybrid points assistance and outperforms state-of-the-art approaches.

FPGA implementation of a multi-channel pipelined large FFT architecture is challenging due to its complex inter-channel data scheduling, high-throughput requirement, and resource-constrained hardware. By transforming to 2D-FFT implementation, investigating different binary tree schemes, and exploring various radices, butterflies, as well as data path structures, many hardware architectures have been designed to enhance single-channel large FFT or multi-channel medium-small size FFT performance. These designs fall short in addressing the demands of multi-channel pipelined and large FFT applications. In this article, a self-attention multipath delay feedback (SA-MDF) algorithm is proposed to analyze and identify the most critical bottleneck, then automatically pay attention to improve it, and finally generate the optimal FFT framework by exhaustively exploring the design space. The proposed algorithm alleviates the design difficulties and speeds up the FPGA implementation. Furthermore, an approximate roofline model and a novel binary tree scheme are introduced to further minimize the utilization of on-chip memory. A comprehensive comparison in terms of principles, implementation methods, and optimization effects is conducted when compared with other multi-channel FFT architectures. Experimental results show that the proposed FFT architectures are superior to other FFT implementations in terms of channel count, FFT length, high-throughput data arrangement, and adaptability to diverse hardware platforms.

Millimeter-wave (mm-Wave) transceivers with high power, enhanced peak/back-off efficiency, and low noise are highly desired in 5G New Radio (NR) systems. This article presents a 26.5–29.5-GHz Doherty power amplifier (PA)-low noise amplifier (LNA) with a compact transformer-based matching network (MN). The boundary conditions of a three-port Doherty PA-LNA combiner network are thoroughly analyzed. Building on this analysis, a novel transformer-based MN topology is proposed, enabling simultaneous realization of Doherty operations, noise matching, and high isolation within a time-division duplex (TDD) transmit/receive (T/R) front-end. To validate this concept, a Doherty PA-LNA prototype is designed and fabricated in 65-nm bulk CMOS, occupying a core area of$1.3\times 0.65$mm2. In TX mode, it achieves a maximum output 1-dB gain compression point (OP1dB) of 17.5-dBm with a power-added efficiency (PAE) of 14.6% at 6-dB power back-off (PBO). The drain efficiency (DE) exhibits a superior 1.66 times efficiency enhancement over the normalized class-B DE at 6-dB PBO. In RX mode, it achieves a minimum noise figure (NF) of 4.8 dB at 28.6 GHz with an in-band NF ranging from 4.8 to 5.4 dB. The chip consumes 255 mW at OP1dB in TX mode and 33 mW in RX mode.

This paper presents a K-band eight-element dual-beam receiver in 65-nm CMOS for satellite communications. Current-sharing low-noise amplifiers (LNAs) and passive beamforming networks (BFNs) are employed to achieve low-power operation. Single power supply and daisy-chain digital-control scheme could simplify the phased array design. On-chip low dropout regulator (LDO) reduces the sensitivity of the channel gain to the supply-voltage variation by keeping a steady bias for the amplifiers. The chip utilizes a flip-chip chip-scale package (FCCSP) and occupies 5.2 × 6.4 mm2. Each element achieves an electronic gain of 25 dB, a noise figure (NF) of 2.7 dB, and an input-referred 1-dB gain compression point of -37 dBm at 19.5 GHz. Each element achieves a 360° phase shifting range with a 5.625° resolution and a 15.5-dB attenuation range with a 0.5-dB step. The total power consumption is 275.3 mW (corresponding to 17.2 mW per element per beam). The proposed receiver demonstrates the state-of-the-art NF among K-band beamformers in bulk CMOS processes.

The atomic norm minimization (ANM) is widely applied in the off-grid direction of arrival (DOA) estimation as a super-resolution method. However, the workload of ANM will become unbearable with the rapid growth of snapshots. At the same time, the follow-up work for existing ANM algorithms is quite cumbersome. Therefore, in this paper, we propose an improved off-grid DOA estimation algorithm based on expanded matrix and matrix pencil. The contribution is as follows: Firstly, the cross-covariance matrix (CCM) is expanded to simplify the operation of ANM while maintaining high resolution. Secondly, weighted matrix enhancement matrix pencil (WMEMP) is introduced by us. Computing efficiency is significantly improved by the structure of WMEMP. Optimal weight also improves resolution. The results show that our algorithm is suitable for L-shaped arrays, with good robustness, similar accuracy to the ANM algorithm, and higher computing efficiency compared with other off-grid estimation algorithms.

Random multiple access is a candidate for next-generation multiple access because of its flexibility and simplicity. In this paper, we propose a novel random access scheme, each user holds a signature code as its ID decides whether to transmit or retransmit their packet based on its state and network status, while the receiver chooses to receive and provide feedback to the corresponding user based on the user’s state. We model the system as a Markov chain using a stochastic game, analyze its steady-state distribution, and theoretically analyze the performance of the network. In addition, we analyzed the optimal signature code capability. The simulation results match well with the theoretical analysis, and the proposed protocol significantly outperforms the existing protocols.

This letter examines the fusion detection challenge in multi-frequency-based passive bistatic radar (MF-PBR) in the presence of the signal-to-noise ratio (SNR) difference. The traditional equal-gain-combination (EGC) based methods suffer from performance loss when the SNR difference over the multiple frequency channels is enormous. To this end, a novel weighted combination strategy is proposed for quasi-coherent MF integration. Specifically, a closed-form expression of the fusion weight is derived via modifying the non-coherent integration-based fusion weight with respect to the MF channel parameters. Moreover, for the implementation of the proposed fusion technique, an iterative algorithm for the estimation of MF channel parameters is also developed. Field test results verify the superiority of the proposed method in terms of detection capability.

To address the issue that existing multi-domain fusion methods do not consider data correlation within mini-batch data, the first attempt is made in this letter to propose a method based on sample inter-correlation learning multi-domain fusion network (SIMFNet), which aims to accommodate multivariate domain data and further enhance the aquatic human activity recognition performance. To fully utilize the radar multidimensional information, the three-branch convolution neural network (CNN) feature extractor is first employed to extract domain-specific features from the time-range map (TRM), time-Doppler map (TDM) and cadence velocity diagram (CVD). Then, the multi-domain features are fused and fed into the graph construction layer (GCL) to generate instance graphs. Next, a graph aggregation layer (GAL) is applied to aggregate node information from various-hop neighborhood domains. Finally, node-level classification is used to achieve aquatic human activity recognition. The experimental results evaluated on the built aquatic human activity recognition dataset demonstrate that the proposed SIMFNet has better generalization performance than the state-of-the-art multi-domain fusion methods.

Deep learning (DL) frameworks are widely used in various applications due to their superiority over conventional handcrafted feature-based algorithms. However, applying DL to time-range feature map-based people flow detection (PFD) with frequency modulated continuous wave (FMCW) radar still faces several challenges: 1) simultaneously achieving people counting and motion direction recognition requires a unified framework, 2) existing mainstream network backbones designed for semantic information-rich optical images or natural language suffer performance loss in time-range feature maps with weak semantic information, and 3) the construction of labeled data sets in PFD scenes is usually costly, and limited data lead to performance loss due to overfitting. Therefore, this paper proposes novel solutions from various aspects to efficiently apply DL to time-range feature map-based PFD. First, new preprocessing pipelines with Doppler spectrum analysis-based feature map truncation are proposed for the first time to simultaneously achieve people counting and direction recognition using a single radar in the radar PFD field. Second, a novel lightweight multiscale feature space fusion-based convolutional neural network backbone (MFSNet) is designed to efficiently extract multichannel differentiated representative features from time-range feature maps. Finally, a multiencoder-classifier cotraining architecture based on embedded features with two data synthesis methods and a newly designed loss function is proposed to improve the generalization ability of the algorithm. Using the test data set collected from real scenes, the performance comparison results show that the proposed PFD algorithm outperforms the state-of-the-art algorithms and ablation studies demonstrate the effectiveness of each component of the proposed algorithm in PFD.

Inspired by animals’ long-distance migration behaviors, we proposed a novel and reliable long-distance true navigation method based on the measurement of geomagnetic fields. By establishing a two-dimensional (2D) gradient approaching coordinate plane with geomagnetic intensity and inclination, long-distance true navigation can be achieved from any starting spots in this area. Without other calibration and auxiliary information, this navigation method is highly independent, which can be used in electromagnetic shielding circumstances, such as deep ocean. The geomagnetic intensity and inclination gradients are estimated with real-time measurement of the geomagnetic field along the navigation trajectory. With estimated local gradients, 2D gradient approaching algorithm can be applied to predict the heading direction, and the agent can press on towards the destination step by step. Theoretical analysis and Monte Carlo simulations verified the feasibility and practicality of this method, proving its robustness in the presence of interference and different measurements errors. It can be a promising candidate in the design of navigation systems used in autonomous underwater vehicles or platform.

Radar is widely used in human activity recognition because of its powerful micro-doppler feature capture capability and environmental adaptability. In this work, we propose a novel radar-based potential drowning detection system. To enhance the cross-domain fusion efficiency and intra-domain feature learning, we design a two-stage fusion network for the drowning detection system. In the first-stage fusion, we integrate the encoded features of three-domain radar maps along either the temporal or spatial dimension. In the second-stage fusion, we use Attention-LSTM and 1D-CNN to extract deep information from temporal-fused and spatial-fused features, and further combine these features using a trainable weighted average strategy. Based on our proposed novel fusion architecture, fine-grained aquatic human activity recognition is achieved. In the experiments, we collect a nine-class aquatic human activity dataset. The experimental results demonstrate the superiority of the proposed TSFNet over the state-of-the-art models. The dataset and the associated codes are available at: https://github.com/DingdongD/aquatic-activity-dataset.

The multi-satellite tracking (MST) for fast link switch is of great importance in guaranteeing the link quality of low earth orbit (LEO) satellite communications. In the marine satellite communication scenario, the position and attitude of the ship and the satellite are known. However, due to the noise from uninterested flight vehicles and the communication path, as well as the effect of the coupling motion between the LEO satellite and the ship, an MST system is necessary to achieve high tracking accuracy. In this work, we propose an MST algorithm for shipborne digital phased arrays, which exploits the constellation orbit characteristics and then associates the interested target states with the orbit parameters. Upon performing a coarse MST using the probabilistic multi-hypothesis tracking (PMHT), the proposed algorithm further picks interested targets and updates their states by the satellite orbit association (SOA). In the process, states of the interested targets are determined according to the orbit parameters in a designed linear constellation orbit model (LCOM), and new strategies are applied to construct and update the LCOM to ensure its effectiveness. Numerical simulations show that the proposed algorithm performs better over the conventional ones.

Estimating ocean wave height is vital for coastal activities. Herein, we propose a method for the estimation of wave height using a phase time series of the frequency-modulated continuous-wave (FMCW) millimeter-wave (MMW) radar. We simulated a one-dimensional wavy ocean surface under different wind speeds and the electromagnetic echo of MMW radar on the ocean surface. The simulation results show that the unwrapping phase time series-based wave heights are highly consistent with the actual wave heights. This shows that the phase time series of the FMCW MMW radar is effective for estimating wave height, and the FMCW MMW radar can be used to monitor ocean states.

Multitarget detection is very challenging, especially when targets are densely distributed. In conventional constant false alarm rate (CFAR) detection algorithms, the detection threshold is determined based on a pre-estimated background level (BL). However, interfering targets inevitably lead to inaccurate BL estimation, resulting in degraded detection performance. In our previous work, a new solution was developed and verified to be effective; in that solution, compressed sensing (CS)-based detector performs target detection without relying on BL estimation, and a subsequent CFAR regulation processor achieves the desired false alarm rate based on the statistical information of the reduced samples obtained by removing the detected targets and adjacent guard cells from the original samples. However, for scenes with a high target density, the CS-based detector suffers from performance degradation while acquiring the correlation between linear measurements of the signal and the sensing matrix due to the high local signal sparsity. To address this shortcoming, this article proposes a deep neural network (DNN)-based detector that further improves the detection performance by converting the target detection into a problem of peak sequence classification of frequency intensity (FI) measurements from radar. A DNN detector trained on augmented simulated data shows excellent generalization ability for deployment in real scenes. In addition, to achieve better computational efficiency, a Taylor-series-based approximate maximum likelihood estimator with an explicit expression is applied in the CFAR regulation process for the first time. Both simulation and field tests are performed to verify the superiority of the proposed algorithm over conventional algorithms.

Particle swarm optimization (PSO) has received great attention for its ftexibility and brevity in conformal antenna array pattern synthesis (CAAPS). However, it also suffers from local optimum and increased computational complexity. In this paper, a solution space pruning particle swarm optimization (SSP-PSO) dedicated to the CAAPS is proposed, which selectively optimizes the peak sidelobe level (PSLL) among the multiple optimization objectives and accordingly applies iterative fast Fourier transform (IFT) at the first stage, and then avoids ineffective search by accomplishing SSP on the basis of dynamic range ratio constraints and the element excitation that is obtained from the foregoing PSLL optimization. The simulation results verify the superiority of the proposed method in terms of both convergence accuracy and convergence speed.

Constant false alarm rate (CFAR) detection algorithms, which are widely used in frequency-modulated continuous wave (FMCW) radar systems, achieve target detection by employing a threshold determined on the basis of a predicted background level. However, in multitarget scenarios, the multitarget shadowing effect can lead to inaccurate prediction of the background level and improper setting of the threshold, which then results in severely degraded CFAR performance. To combat this multitarget shadowing effect, a novel CFAR algorithm based on sparsity adaptive correlation maximization (SACM-CFAR) is proposed in this work. The proposed SACM-CFAR algorithm realizes target detection by utilizing the correlation between linear measurements of the radar intermediate frequency (IF) signal and the sensing matrix. To achieve a desired false alarm rate, the proposed algorithm determines the threshold by estimating the distributed parameters of the reduced sample set obtained by removing the detected targets from the original sample set. Both simulation results and field test results verify that the proposed algorithm outperforms conventional algorithms in multitarget scenarios.

Constant false alarm rate (CFAR) detection plays a vital role in frequency modulated continuous wave (FMCW) radar systems. Most existing CFAR algorithms need to estimate background level before determining the detection threshold. In a multi-target scene, existence of interference targets could cause inaccurate estimation of background level, which will then lead to severely degraded performance of the CFAR algorithm. This is also called multi-target shadowing effect. To overcome the challenge, a compressed sensing CFAR (CS-CFAR) is proposed in this work, which detects the target of interest (TOI) relying on the correlation between the linear measurements of signal and the sensing matrix, instead of relying on an estimated background level as in conventional CFAR algorithms, and achieves a required false alarm rate by estimating the distributed parameters of reduced sample model. Through doing so the proposed CS-CFAR algorithm effectively mitigates impact of interference targets on the detection performance. Superiority of the proposed algorithm over the conventional ones in multi-target scenarios is verified by both simulation and test results.

The fundamental multidimensional line spectral estimation problem is addressed utilizing the Bayesian methods. Motivated by the recently proposed variational line spectral estimation (VALSE) algorithm, multidimensional VALSE (MDVALSE) is developed. MDVALSE inherits the advantages of VALSE such as automatically estimating the model order, noise variance and providing uncertain degrees of frequency estimates. Compared to VALSE, the multidimensional frequencies of a single component is treated as a whole, and the probability density function (PDF) is projected as independent univariate von Mises distribution to perform tractable inference. Besides, for the initialization, efficient fast Fourier transform (FFT) is adopted to approximate the marginal posterior PDF of frequencies. Numerical results demonstrate the effectiveness of the MDVALSE, compared to state-of-art methods.

In this letter, the problem of sparse signal reconstruction from quantized noisy magnitudes y = Q (|z + w| + n) is studied, where z = Ax, Q(·) denotes a quantizer, and x is the sparse signal. According to expectation propagation, the abovementioned problem can be solved by exchanging extrinsic information between the standard linear model (SLM) module and the minimum mean square error (MMSE) module. For the MMSE module, exploiting the fact that the likelihood is only a function of the absolute value of z, the magnitude and phase information of z are decoupled, and only the posterior means and variances of the magnitude of z are calculated. While for the SLM module, the approximate message passing (AMP) is used. To obtain the closed-form expression during the iteration, a novel approximation is adopted. We refer to the resulting algorithm as generalized AMP (Gr-AMP) based quantized phase retrieval algorithm. Finally, several numerical simulations are conducted to demonstrate the performances of the proposed approach.

The problem of designing waveform and quantisation thresholds is studied in a multistatic radar setting, where distributed receivers are connected to a fusion centre via capacity constraints. Different from the previous cloud radio‐multistatic radar system which utilises an additive quantisation Gaussian noise model, a real quantisation system is designed. The authors first optimise the waveform without quantisation. Then they compress the received signal at receivers into a scalar without any performance degradation. Furthermore, the scalar quantiser is adopted and quantisation thresholds are designed. Numerical simulations are performed and the effectiveness of combined waveform and thresholds optimisation strategy is demonstrated.

In this study, the problem of recovering structured sparse signals with a priori distribution whose structure patterns are unknown is studied from one‐bit adaptive (AD) quantised measurements. A generalised approximate message passing (GAMP) algorithm is utilised, and an expectation maximisation (EM) method is embedded in the algorithm to iteratively estimate the unknown a priori distribution. In addition, the nearest neighbour sparsity pattern learning (NNSPL) method is adopted to further improve the recovery performance of the structured sparse signals. Numerical results demonstrate the effectiveness of GAMP‐EM‐AD‐NNSPL method with both simulated and real data.

The impact of random compression on the Fisher information matrix (FIM) and the Cramér-Rao bound (CRB) is studied when estimating unknown complex parameters in the perturbed linear model. A random compression matrix is considered whose elements are i.i.d. standard complex normal random variables. The FIM averaged over compression is equal to a scalar of the FIM before compression plus an additional term. The upper and lower bounds of the CRB averaged over the random compression matrix are also given. Finally, numerical results are conducted to verify our theoretical results.

Passive radar has shown lots of advantages as compared to active radar and has become a significant branch of bistatic radar. Detection difficulty of non-cooperative reflected signals at very low SNR is a well-known challenge in passive radar, and previous works have mainly studied multi-user co-operative detection based solutions. In this work, we study a passive radar system using the television satellite as illuminator of opportunity(IO), and focus on finding solution for detection challenges caused by time-varying interference and high-speed moving target through improving performance of each single detector. A multi-channel based passive detection strategy is proposed. The proposed method is characterized by sufficient information based fusion and weighted combining. Theoretical analysis and Monte Carlo simulation are conducted to evaluate the performance of our strategy.