Signal Model Research Articles

Height Measurement Method for Meter-Wave Multiple Input Multiple Output Radar Based on Transmitted Signals and Receive Filter Design

To address the issue of low-elevation target height measurement in the Multiple Input Multiple Output (MIMO) radar, this paper proposes a height measurement method for meter-wave MIMO radar based on transmitted signals and receive filter design, integrating beamforming technology and cognitive processing methods. According to the characteristics of beamforming technology forming nulls at interference locations, we assume that the direct wave and reflected wave act as interference signals and hypothesize a direction for a hypothetical target. Then, the data received are processed to obtain the height of low-elevation-angle targets using a cognitive approach that jointly optimizes the transmitted signal and receive filter. Firstly, a signal model for the meter-wave MIMO radar based on the transmit weight matrix is established under low-elevation scenarios. Secondly, the signal model is analyzed and transformed. Thirdly, the beamforming algorithm that jointly optimizes the transmitted signals and receive filter is derived and analyzed. The algorithm maximizes the output Signal-to-Interference-plus-Noise ratio (SINR) of the receiver by designing the transmit weight matrix and receive filter. The optimization problem based on the SINR criterion is non-convex and difficult to solve. We transformed it into two sub-optimization problems and approximated the optimal solution through an alternating iteration algorithm. Finally, the proposed height measurement algorithm is compared with the Generalized Multiple Signal Classification (GMUSIC) and Maximum Likelihood (ML) height measurement algorithms. Simulation results show that the proposed algorithm can realize the height measurement of low-elevation targets. Compared to the GMUSIC and ML algorithms, it demonstrates superior performance in terms of computational complexity and multi-target elevation estimation.

Digitizing Low-Frequency Analog Control Circuit Using Bilinear Function Algorithm

In the past, low-frequency analog control circuits were largely used in analog control systems. With the rapid development of modern digital electronic technology, the digitization of traditional low-frequency analog control circuits while retaining the same functions as the original analog control systems has become an important trend in the field of electronic design. For this reason, this study aimed to develop an analog signal digitization model for control signals at a low frequency below 10 Hz. In terms of signal receiving and transmission requirements, the proposed model is configured with analog-to-digital converter (ADC), digital-to-analog converter (DAC), and pulse width modulation (PWM) functions. In the development environment, the input and output signals are first normalized and processed with PWM, enabling the digital signal processor to deal with analog signal receiving, processing, and external transmission. To replace the existing compensator in analog circuits, a digital compensator is used in the digital signal processor. Based on the bilinear function in MATLAB software, the parameter values demanded for the digital compensator can thus be obtained, achieving the automatic calculation of bilinear transformation in the system. Multisim simulation is then used to simulate analog circuit systems for comparison with the digitized outcomes. The experimental results reveal that the performance of the designed digital control circuit matches the simulation outcomes in both Bode gain (dB) and phase responses when the signal frequency is below 10 Hz. The effectiveness of the digitized analog control circuit for low-frequency control signals is therefore confirmed.

A Mathematical Approach to the Stability of Electroencephalography Signals Model Using Integral Equations During Epileptic Seizures

In this study, the mathematical stability of EEG signal models represented by the integral equation during an epileptic seizure is investigated. The modelling of the EEG signals should be done with utmost accuracy to understand the actual nature of the brain dynamics and to enhance seizure prediction approaches. We investigate the stability of such models by applying Lyapunov's stability theorem and show that stability is powerfully dependent on parameters such as seizure duration and intensity, along with neural connectivity. We demonstrate that some of the parameter regimes yield stable behaviour, while others give rise to instability, making EEG interpretation complicated during epileptic seizures. By comparing integral equation-based models with current approaches, their advantages are highlighted in capturing the transient dynamics of epileptic activity. This work focuses on developing solid mathematical frameworks that will allow for real-time EEG monitoring. It contributes to the wider knowledge on the dynamics of seizures in the brain and will have practical implications for the development of neurological diagnostic and therapeutic tools.

Converters with Reduced Duty Cycle Range and Ground Connected to the Positive Input Voltage Terminal

A family of converters with reduced duty cycle range is treated here. Two-step-down, two step-up, and four step-up-down converters are investigated. Four different voltage transformation ratios can be found. For all converters, the large signal and the small signal models of the ideal case are given. Furthermore, the connections of the operating point values and the voltage stress across the semiconductors are indicated. Simulations for the steady-state are shown. A method to calculate the transfer functions is proposed and one transfer function for type 1 is shown as an example. The inrush current, when the converter is connected to a stiff input voltage source, is studied with the help of calculations, and time- and phase-diagrams. Several variations and combinations are also shown.

Emotion Recognition Model of EEG Signals Based on Double Attention Mechanism.

Emotions play a crucial role in people's lives, profoundly affecting their cognition, decision-making, and interpersonal communication. Emotion recognition based on brain signals has become a significant challenge in the fields of affective computing and human-computer interaction. Addressing the issue of inaccurate feature extraction and low accuracy of existing deep learning models in emotion recognition, this paper proposes a multi-channel automatic classification model for emotion EEG signals named DACB, which is based on dual attention mechanisms, convolutional neural networks, and bidirectional long short-term memory networks. DACB extracts features in both temporal and spatial dimensions, incorporating not only convolutional neural networks but also SE attention mechanism modules for learning the importance of different channel features, thereby enhancing the network's performance. DACB also introduces dot product attention mechanisms to learn the importance of spatial and temporal features, effectively improving the model's accuracy. The accuracy of this method in single-shot validation tests on the SEED-IV and DREAMER (Valence-Arousal-Dominance three-classification) datasets is 99.96% and 87.52%, 90.06%, and 89.05%, respectively. In 10-fold cross-validation tests, the accuracy is 99.73% and 84.26%, 85.40%, and 85.02%, outperforming other models. This demonstrates that the DACB model achieves high accuracy in emotion classification tasks, demonstrating outstanding performance and generalization ability and providing new directions for future research in EEG signal recognition.

A Motion Compensation Method for Terahertz SAR Imaging with a Large Squint

Terahertz-band squint synthetic aperture radars (SARs) can obtain high-resolution images and have application potential in airborne radar systems. However, airborne radars usually have a large squint, which has led to traditional SAR algorithms no longer being applicable to airborne SARs. Additionally, terahertz radar imaging systems are more susceptible to the error induced by the platform’s motion. This paper proposes a motion compensation method for terahertz SAR imaging with a large squint angle. First, the signal model of motion compensation is derived, and the processing flow of imaging and motion compensation is detailed. Second, some simulations and experiments are conducted, and the results are reported. The results indicate that the proposed method can effectively correct the motion errors, and the signal model and processing flow are verified.

EPILEPTIC SEIZURE DETECTION FROM EEG SIGNALS WITH RECURRENT NEURAL NETWORKS BASED CLASSIFICATION MODEL

Epileptic seizures are a neurological disorder that occurs as a result of sudden and uncontrolled electrical activities of the non-contagious brain. This condition may cause the person to lose normal activities temporarily. Epileptic seizures are a severe disease that affects approximately 60 million people in the world, usually manifested by symptoms such as loss of consciousness, muscle twitching, sudden sensory changes, or behavioural changes [1]. Genetics, brain injury, hormonal fluctuations, infections, or metabolic problems are some of the possible causes of epileptic seizures. Although the severity and duration of the seizure varies from person to person, it is usually very short and rarely reaches a point where it endangers human life. However, such seizures need to be recognized as soon as possible in order to improve the quality of life of individuals and reduce the frequency of seizures. Epileptic seizures are a manageable disease with early diagnosis and appropriate treatment. Recognizing epileptic seizures begins with understanding a person's symptoms and triggering factors. These symptoms may include loss of consciousness, muscle twitches, sudden sensory changes, and behavioural changes. The symptoms of seizures, past medical history, and neurological examinations are essential in the diagnosis process. From past to present, many methods have been developed for the early diagnosis and detection of epileptic seizures [2]. One of these is analyzing the brain's neural activities using electroencephalography (EEG), which helps experts make a diagnosis. Although EEG signals are used as a powerful tool in epileptic seizure recognition, distinguishing the signals within them is both costly and requires highly expert experience. Therefore, this study proposed an automatic classification model for pre-processed EEG signals using Dual-Tree Complex Wavelet Transform (DT-CWT) based on deep learning-based Recurrent Neural Networks (RNN) architecture to assist experts. Compared to classical machine learning methods, deep learning-based models require less manual feature engineering because they perform data automatically thanks to deep networks instead of manually selecting and transforming the data features. These advantages make the model more general and flexible. The proposed model aims to classify EEG signals and detect epileptic seizures effectively and quickly in the early stages.

Model-based CBCT scatter correction with dual-layer flat-panel detector.

Recently, the popularity of dual-layer flat-panel detector (DL-FPD) based dual-energy cone-beam CT (CBCT) imaging has been increasing. However, the image quality of dual-energy CBCT remains constrained by the Compton scattered x-rayphotons. The objective of this study is to develop a novel scatter correction method, named e-Grid, for DL-FPD based CBCTimaging. In DL-FPD, a certain portion of the x-ray photons (mainly low-energy [LE] primary and scattered photons) passing through the object are captured by the top detector layer, while the remaining x-ray photons (mainly high-energy [HE] primary and scattered photons) are collected by the bottom detector layer. A linear signal model was approximated between the HE primary and scatter signals and the LE primary and scatter signals. Physical calibration experiments were performed on cone beam and fan beam to validate the aforementioned signal model via linear fittings. Monte Carlo (MC) simulations of a 10 cm diameter water phantom were conducted on GATE at first to verify this newly developed scatter estimation method. In addition, physical validation experiments of water phantom, head phantom, and abdominal phantom were carried out on a DL-FPD based benchtop CBCT imaging system. The image non-uniformity (NU), which represents the relative difference between the center and the edges of CT images, was measured to quantify the reduction of image shading artifacts. Finally, multi-material decomposition wasconducted. The MC results, CBCT images and line profiles, showed that the newly proposed e-Grid approach was able to accurately predict the scatter distributions in both shape and intensity. As a result, uniform CBCT images that are close to the scatter artifact-free reference images can be obtained. Moreover, the physical experiments demonstrated that the e-Grid method can greatly reduce the shading artifacts in both LE and HE CBCT images acquired from DL-FPD. Results also demonstrated that the e-Grid method is effective for varied objects that having different diameters (from 10 to 28 cm). Quantitatively, the NU value was reduced by over 77% in the LE CBCT image and by over 66% in the HE CBCT image on average. As a consequence, the accuracy of the decomposed multi-material bases, iodine and gadolinium, was substantially improved. The Compton scattered x-ray signals could be significantly reduced using the proposed e-Grid method for DL-FPD based dual-energy CBCT imagingsystems.

Concept and theoretical performance analysis for decentralised digital synchronisation in distributed radar sensor networks

AbstractThis paper presents a decentralised technique for achieving frequency, time, and phase synchronisation of platforms cooperating in a distributed radar sensor network. The proposed method is advantageous for existing digital radar systems as it can be implemented entirely in the software without the use of additional RF hardware required by other techniques. The synchronisation signal model for signals transmitted and received in various clock domains is presented and an estimation model is subsequently derived for estimating and correcting the clock drifts. A modified version of a previously developed phase and clock bias correction procedure is outlined for correcting time and phase after frequency synchronisation. A comprehensive theoretical performance analysis of the technique is performed in which the expected maximum achievable performance is derived in terms of the Cramér–Rao lower bound for frequency, time, and phase measurements. Multiple Monte Carlo simulations show that the proposed technique approaches this performance limit. Finally, a simulated distributed transmit beamforming scenario is provided to show the application of the proposed technique in a practical system architecture. The results of this show that as the signal‐to‐noise ratio approaches moderate levels, the proposed synchronisation technique enables the beamforming network to achieve nearly optimal coherent energy gain at the beamforming destination.

Detecting flying objects in synthetic aperture radar images using Moving Target Indicator methods

AbstractThe growing proliferation of synthetic aperture radar (SAR) sensors brings the tantalising prospect of extending their utility into ‘novel’ applications. One potential extension is the detection of fast moving and accelerating flying objects in SAR imagery. However, since SAR image formation typically assumes the scene to be static over the coherent processing interval, moving objects give rise to blurred point spread functions, significant range migration and even potential aliasing of target signatures. The result is reduced target to clutter ratio (TCR) and poor detection performance. Successful detection of airborne targets thus requires compensation for potentially large target acceleration and velocity values observed over the comparatively long dwell times typical of practical SAR collection paradigms. This paper considers this problem and presents two main ideas to achieve this goal: a carefully constructed Moving Target Indicator (MTI) detection method implemented using real‐world Ingara SAR data, and a theoretical ground clutter suppression method. The MTI detection method combines several well‐known techniques for the flying target detection problem: interferometric processing, clutter suppression, and autofocus, and provides an extended acceleration phase compensation technique for highly accelerating targets such as planes. This proposed processing pipeline has been applied to experimental data of a plane during take off (a challenging Doppler unambiguous moving target), with the goal of continued detecting and tracking of this target. A generalised SAR signal model is presented that parameterises a flying moving target signature in terms of range and azimuthal target velocities and accelerations. Data driven approaches for estimating these motion parameters are examined and applied to experimental data acquired with the Ingara SAR sensor. The detection method was found to improve TCR by around 6 dB, along with superior detection and tracking performance. Following this, a theoretical study into suppressing ground clutter via multi‐channel cross‐track interferometry is investigated. Three separate ground clutter suppression methods, coherent subtraction, conventional beamforming, and minimum variance distortionless response (MVDR) beamformer, are presented then analysed using stochastic simulations. The MVDR adaptive beamformer method was found to provide the best performance for the scenario simulated.

Processing piano audio: research on an automatic transcription model for sound signals

Automatic transcription of sound signals can convert audio to musical notes, which has significant research value. This paper extracted dual-channel constant Q transform (CQT) spectra from piano audio as features. In the design of the automatic transcription model, a CNN was employed to extract local features and then combined with a Transformer model to obtain global features. A CNN-Transformer automatic transcription model was established using a two-layer CNN and three-layer Transformers. Experiments were conducted on the MAPS and MAESTRO datasets. The results showed that dual-channel CQT outperformed short-time Fourier transform (STFT) and mono CQT in auto-transcription. Dual-channel CQT achieved the best results on frame-level transcription for the MAPS dataset, with a P value of 0.9115, an R value of 0.8055, and an F1 value of 0.8551. A sliding window with seven frames yielded the best transcription results. Compared with the deep neural network and CNN models, the CNN-Transformer model demonstrated superior performance, achieving an F1 value of 0.8551 and 0.9042 at the frame level for MAPS and MAESTRO datasets, respectively. These findings confirm the designed model's reliability for automatic piano audio transcription and highlight its practical applicability.

Two-neutrino double electron capture of 124Xe in the first LUX-ZEPLIN exposure

Abstract The broad physics reach of the LUX-ZEPLIN (LZ) experiment covers rare phenomena beyond the direct detection of dark matter. We report precise measurements of the extremely rare decay of 124Xe through the process of two-neutrino double electron capture, utilizing a 1.39 kg × yr isotopic exposure from the first LZ science run. A half-life of T 1 / 2 2 ν 2 EC = ( 1.09 ± 0.1 4 stat ± 0.0 5 sys ) × 1 0 22 yr is observed with a statistical significance of 8.3σ, in agreement with literature. First empirical measurements of the KK capture fraction relative to other K-shell modes were conducted, and demonstrate consistency with respect to recent signal models at the 1.4σ level.

Estimation of fatty acid composition in mammary adipose tissue using deep neural network with unsupervised training.

To develop a deep learning-based method for robust and rapid estimation of the fatty acid composition (FAC) in mammary adipose tissue. A physics-based unsupervised deep learning network for estimation of fatty acid composition-network (FAC-Net) is proposed to estimate the number of double bonds and number of methylene-interrupted double bonds from multi-echo bipolar gradient-echo data, which are subsequently converted to saturated, mono-unsaturated, and poly-unsaturated fatty acids. The loss function was based on a 10 fat peak signal model. The proposed network was tested with a phantom containing eight oils with different FAC and on post-menopausal women scanned using a whole-body 3T MRI system between February 2022 and January 2024. The post-menopausal women included a control group (n = 8) with average risk for breast cancer and a cancer group (n = 7) with biopsy-proven breast cancer. The FAC values of eight oils in the phantom showed strong correlations between the measured and reference values (R2 > 0.9 except chain length). The FAC values measured from scan and rescan data of the control group showed no significant difference between the two scans. The FAC measurements of the cancer group conducted before contrast and after contrast showed a significant difference in saturated fatty acid and mono-unsaturated fatty acid. The cancer group has higher saturated fatty acid than the control group, although not statistically significant. The results in this study suggest that the proposed FAC-Net can be used to measure the FAC of mammary adipose tissue from gradient-echo MRI data of the breast.

Reproducibility of intravoxel incoherent motion quantification in the liver across field strengths and gradient hardware.

To evaluate reproducibility and interlobar agreement of intravoxel incoherent motion (IVIM) quantification in the liver across field strengths and MR scanners with different gradient hardware. Cramer-Rao lower bound optimization was performed to determine optimized monopolar and motion-robust 2D (b-value and first-order motion moment [M1]) IVIM-DWI acquisitions. Eleven healthy volunteers underwent diffusion MRI of the liver, where each optimized acquisition was obtained five times across three MRI scanners. For each data set, IVIM estimates (diffusion coefficient (D), pseudo-diffusion coefficients ( and ), blood velocity SDs (Vb1 and Vb2), and perfusion fractions [f1 and f2]) were obtained in the right and left liver lobes using two signal models (pseudo-diffusion and M1-dependent physical) with and without T2 correction (fc1 and fc2) and three fitting techniques (tri-exponential region of interest-based full and segmented fitting and blood velocity SD distribution fitting). Reproducibility and interlobar agreement were compared across methods using within-subject and pairwise coefficients of variation (CVw and CVp), paired sample t-tests, and Bland-Altman analysis. Using a combination of motion-robust 2D (b-M1) data acquisition, M1-dependent physical signal modeling with T2 correction, and blood velocity SD distribution fitting, multiscanner reproducibility with median CVw = 5.09%, 11.3%, 9.20%, 14.2%, and 12.6% for D, Vb1, Vb2, fc1, and fc2, respectively, and interlobar agreement with CVp = 8.14%, 11.9%, 8.50%, 49.9%, and 42.0%, respectively, was achieved. Recently proposed advanced IVIM acquisition, signal modeling, and fitting techniques may facilitate reproducible IVIM quantification in the liver, as needed for establishment of IVIM-based quantitative biomarkers for detection, staging, and treatment monitoring of diseases.

Revised bounds on local cosmic strings from NANOGrav observations

In a recent paper, the NANOGrav collaboration studied new physics explanations of the observed pulsar timing residuals consistent with a stochastic gravitational wave background (SGWB) [1], including cosmic strings in the Nambu-Goto (NG) approximation. Analysing one of current models for the loop distribution, it was found that the cosmic string model is disfavored compared to other sources, for example, super massive black hole binaries (SMBHBs). When both SMBHB and cosmic string models are included in the analysis, an upper bound on a string tension Gμ≲ 10-10 was derived. However, the analysis did not accommodate results from cosmic string simulations in an underlying field theory, which indicate that at most a small fraction of string loops survive long enough to emit GW. Following and extending our previous study [2], we suppose that a fraction f NG of string loops follow NG dynamics and emit only GWs, and study the three different models of the loop distribution discussed in the LIGO-Virgo-KAGRA (LVK) collaboration analyses.We re-analyse the NANOGrav 15yrs data with our signal models by using the NANOGrav ENTERPRISE analysis code via the wrapper PTArcade. We find that loop distributions similar to LVK Model B and C yield higher Bayes factor than Model A analysed in the NANOGrav paper, as they can more easily accommodate a blue-tilted spectrum of the observed amplitude. Furthermore, because of the degeneracy of Gμ and f NG in determining the signal amplitude, our posterior distribution extends to higher values of Gμ, and in some cases the uppermost value of credible intervals is close to the Cosmic Microwave Background limit Gμ≲ 10-7.Hence, in addition to the pulsar timing array data, further information about the fraction of long-lived loops in a cosmic string network is required to constrain the string tension.

A multi-scale model for condition-dependent vibrational signals of a double-clamped microbeam near its post-buckling configuration caused by biological adsorption

A multi-scale model for condition-dependent vibrational signals of a double-clamped microbeam near its post-buckling configuration caused by biological adsorption

Multiscale interstitial fluid computation modeling of cortical bone to characterize the hydromechanical stimulation of lacunar-canalicular network.

Bone tissue is a biological composite material with a complex hierarchical structure that could continuously adjust its internal structure to adapt to the alterations in the external load environment. The fluid flow within bone is the main route of osteocyte metabolism, and the pore pressure as well as the fluid shear stress generated by it are important mechanical stimuli perceived by osteocytes. Owing to the irregular multiscale structure of bone tissue, the fluid stimulation that lacunar-canalicular network (LCN) in different regions of the tissue underwent remained unclear. In this study, we constructed a multiscale conduction model of fluid flow stimulus signals in bone tissue based on the poroelasticity theory. We analyzed the fluid flow behaviors at the macro-scale (whole bone tissue), macro-meso scale (periosteum, interstitial bone, osteon and endosteum), and micro-scale (lacunar-osteocyte-canalicular) levels. We explored how fluid stimulation at the tissue level correlated with that at the cellular level in cortical bone and characterized the distributions of the pore pressure, fluid velocity and fluid shear stress that the osteocytes experienced across the entire tissue structure. The results showed that the initial conditions of intramedullary pressure had a significant impact on the pore pressure of Haversian systems, but had a relatively small influence on the fluid velocity. The osteocyte which were located at different positions in the bone tissue received very distinct fluid stimuli. Osteocytes in the vicinity of the Haversian Canals experienced higher fluid shear stress stimulation. When the permeability of the LCN was within the range from 10-21m2 to 10-18m2, the distribution of pressure, fluid velocity and fluid shear stress within the osteon near the periosteum and endosteum was significantly different from that in other parts of the bone. However, when the permeability was less than 10-22m2, such a difference did not exist. Particularly, the flow velocity at the lacunae was markedly higher than that in the canaliculi. Meanwhile, the pore pressure and fluid shear stress were conspicuously lower than those in the canaliculi. In this study, we considered the interconnections of different biofunctional units at different scales of bone tissue, construct a more complete multiscale model of bone tissue, and propose that osteocytes at different locations receive different fluid stimuli, which provides a reference for a deeper understanding of bone mechanotransduction.

Impact Analysis of Orthogonal Circular-Polarized Interference on GNSS Spatial Anti-Jamming Array

With the continuous advancement of electromagnetic countermeasures, new types of interference signals (e.g., multi-polarization suppression interference) pose a significant threat to conventional Global Navigation Satellite System (GNSS) services, even when the receiver employs a right-handed circularly polarized (RHCP) anti-jamming array. This paper proposes a receiving signal model for orthogonal circularly polarized (OCP) interference signals based on conventional arrays, following an analysis of the non-ideal characteristics of actual arrays. Furthermore, the mechanism by which OCP interference signals affect anti-jamming performance is examined. Power inversion (PI) and linear constrained minimum variance (LCMV) techniques, applied to both uniform linear arrays and central circular arrays, are utilized to verify the impact of these interference signals. Simulation and physical testing demonstrate that OCP interference significantly affects the interference subspace of the conventional RHCP array, potentially leading to a reduction in the anti-jamming performance of the receiver. To effectively suppress multi-polarization interference, anti-jamming GNSS receivers must either ensure the consistency of cross-polarization among the elements of the array or adopt polarization-sensitive arrays.

Проекционный метод определения аномальных датчиков в задачах виброакустики

A common problem in measurements using a set of vibration sensors in vibroacoustics is the lack of an accurate signal model on the receivers. Such uncertainty can lead to processing algorithms errors and an incorrect solution to the inverse problem. Due to a malfunction, vibration sensors data can also be bad, deteriorating the quality of measurements. In practice, the search for such anomalous sensors is a non-trivial task. In this paper, an original method related to projection algorithms is proposed for this problem. It is shown that the proposed method allows detecting anomalous sensors of various types. The method main feature is using both numerical finite element model of the object and experimental data. The efficiency of the proposed method is demonstrated in the localizing vibration sources problem. The new method can be used in other applications where the inverse problem is solved using a numerical or analytical model.

Clutter-Sensing-Driven Space-Time Adaptive Processing Approach for Airborne Sub-Array-Level Digital Array

Sub-array-level digital arrays effectively diminish the computational complexity and sample demand of space-time adaptive processing (STAP), thus finding extensive applications in many airborne platforms. Nonetheless, airborne sub-array-level digital array radar still encounters pronounced performance deterioration in highly heterogeneous clutter environments due to inadequate training samples. To address this issue, a clutter-sensing-driven STAP approach for airborne sub-array-level digital arrays is proposed in this paper. Firstly, we derive a signal model of sub-array-level clutter sensing in detail and then further analyze the influence of the sidelobe characteristics of the conventional sub-array joint beam on clutter sensing. Secondly, a sub-array joint beam optimization model is proposed, which optimizes the sub-array joint beam into a wide beam with flat-top characteristics to improve the clutter-sensing performance in the beam sidelobe region. Finally, we decompose the complex optimization problem into two subproblems and then relax them into the low sidelobe-shaped beam pattern synthesisproblem and second-order cone programming problem, which can be effectively solved. The effectiveness of the proposed approach is validated in a real clutter environment through numerical experiments.