XTXin Tai

Xin Tai

*This is an algorithmically generated author record based on publicly available information.

Xin Tai has a total of 32 co-authors. Here are the top co-authors by number of shared publications:

  • Jinfeng Hu

    School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China

    7 shared publications
  • Kai Zhong

    School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China

    7 shared publications
  • Huiyong Li

    School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China

    5 shared publications
  • Dongxu An

    School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China

    3 shared publications
  • Yiran Zhang

    National Key Laboratory of Digital and Agile Aircrafi Design, AVIC Chengdu Aircraft Design and Research Institute, Chengdu, China

    3 shared publications
  • Yongfeng Zuo

    Science and Technology on Electronic Test and Measurement Laboratory, Ceyear Technologies Company Ltd., Qingdao, China

    3 shared publications

Latest Publications

  • Communication spectrum-compatible MIMO radar: Unimodular waveform design for DOA estimation

    by Jiawei Feng, Jinfeng Hu, Kai Zhong, Xin Tai, Yiran Zhang, Kaizhi Ruan, Jie Wu


  • Ambiguity Function Shaping for Weak Target Detection: A Joint Design Approach for Unimodular Waveform and Mismatched Filter

    by Xiangqing Xiao, Jinfeng Hu, Xin Tai, Yiran Zhang, Yuankai Wang, Jun Liu, Kai Zhong, Huiyong Li

    Unimodular waveform design for ambiguity function (AF) shaping is a key technology in cognitive radar system. Existing methods are limited either by low degrees of freedom (DOF) under matched-filter schemes or by relaxation errors in joint waveform–filter design. To address these issues, we propose a relaxation-free optimization framework with high DOF for unimodular waveform and mismatched filter design, formulated as a nonconvex minimization under constant modulus (CMC) and complex sphere (CSC) constraints. Recognizing that CMC and CSC are naturally satisfied in corresponding Riemannian manifolds, a product manifold for ambiguity function shaping (PMAFS) method is proposed. The proposed approach transforms the problem into an unconstrained optimization over a product manifold and solves it via a parallel conjugate gradient algorithm. An exact-penalty extension, EP-PMAFS, is also developed for precise energy spectral density (ESD) control. Simulation results confirm that the proposed PM optimization framework achieves stronger detection performance for weak target and precise control of the waveform ESD.

  • Binary Waveform Design for Spectrally-Compatible Cognitive MIMO Radar DOA Estimation

    by Kai Zhong, Jinfeng Hu, Dongxu An, Huan Wan, Yiran Zhang, Xin Tai, Yongfeng Zuo, Ye Yuan, Cunhua Pan, Kah Chan Teh, et al.

    Unimodular waveform design is a key technology in cognitive Multiple Input Multiple Output (MIMO) radar systems. Existing research mainly includes two categories: unimodular continuous/binary waveform design for detection, and unimodular continuous waveform design for Direction of Arrival (DOA) estimation without spectral constraints. Different from existing methods, our focus lies in investigating unimodular binary waveform design for DOA estimation within spectrally crowded environments. This problem is formulated as minimizing the mean square error (MSE) for DOA estimation, subject to the constraints of binary waveform and multiple spectral constraints. Due to the spectral constraint and nonconvex nature of the binary waveform constraint, the problem is NP-hard and challenging to solve directly. Fortunately, we observe that the problem can be decomposed into multiple more tractable subproblems by introducing auxiliary variables. Leveraging this characteristic, we propose an efficient Problem Decomposition-based Sequential Optimization (PDSO) method to tackle this problem. The method introduces two auxiliary variables to decompose the problem into two subproblems: one of which can be solved in closed-form, while the other is efficiently addressed by the Binary Alternating Directions Method of Multipliers (B-ADMM) algorithm. Compared to the existing methods, the proposed approach demonstrates superior performance in terms of computational cost, DOA resolution, and suppression of spectral interference.


  • Enhancing Robustness for Polarimetric Radar: Worst-Case SINR Optimization With Constrained Transceiver

    by Kai Zhong, Yipeng Zhong, Jinfeng Hu, Xin Cheng, Huiyong Li, Xu Cheng, Cunhua Pan, Kah Chan Teh, Xin Tai, Xianxiang Yu, et al.

    Polarization introduces an additional dimension to radar signals, providing significant advantages in aerospace applications. Constrained transceiver design for maximizing the Signal-to-Interference-plus-Noise Ratio (SINR) is critical in fully polarimetric radar systems. Existing methods primarily focus on average SINR maximization, which lacks robustness in uncertain regions. To address this, we propose a method for worst-case SINR maximization, aiming to maximize the minimum SINR within uncertain regions, subject to the constant modulus constraint of the waveform and the complex sphere constraint of the receive filter. The problem is non-smooth and non-convex, making existing methods inapplicable. To overcome this, we introduce an auxiliary variable to smooth the problem using the LogSumExp function, which transforms the non-smooth objective into a smooth one by converting the inequality constraints into a continuous, differentiable form. Additionally, we propose the Inequality Constrained Product Manifold Optimization (ICPMO) framework, which guarantees convergence. Experimental results based on measured data show that our method outperforms existing methods in terms of worst-case SINR and target detection probability.


  • Integrated Constrained Manifold Optimization for Wideband OFDM-DFRC Hybrid Beamforming

    by Dongxu An, Jinfeng Hu, Yiran Zhang, Kai Zhong, Xin Tai, Yongfeng Zuo, Huiyong Li, Fulvio Gini, Maria Sabrina Greco

    Hybrid beamforming (HBF) offers a cost effective solution for multiple-input multiple-output dual-functional radar-communication (MIMO-DFRC) systems by reducing hardware complexity and power consumption. While narrowband HBF has been widely studied, wideband HBF design, particularly for transceiver-side hybrid architectures, remains rather underexplored. Existing HBF methods for wideband MIMO-DFRC systems typically adopt a sequential optimization (SO) framework based on the consensus alternating direction method of multipliers (CADMM), assuming a fully digital receiver architecture during the initial optimization stage. This assumption restricts the design flexibility and limits the achievable spectral efficiency (SE) on the communication side. To this end, an integrated constrained manifold optimization (ICMO) framework is proposed. The ICMO method embeds the transmit and receive HBF matrices, subject to constant modulus (CM) and transmit power per subcarrier, into an integrated variable and projects them onto an ICM-space for joint optimization. This transforms the original constrained problem into an unconstrained manifold optimization problem. Subsequently, a parallel simplified quasi-Newton (PSQN) method is developed to optimize the integrated variable within the ICMO framework, ultimately yielding the transmit and receive HBF coding matrices. Simulation results under a Pareto behavior analysis show that the proposed method dominates the CADMM baseline across the SSME and SE frontier, achieving up to $40\%$ higher SE at comparable SSME and about 4.4 dB lower sidelobe levels, thereby enabling spectrum efficient and low sidelobe DFRC design.


  • Movable Antenna Enhanced Beampattern Synthesis for Spectrum Constrained MIMO Radar

    by Dongxu An, Jinfeng Hu, Ren Wang, Xin Tai, Yongfeng Zuo, Kai Zhong, Fulvio Gini, Maria Sabrina Greco

    Designing Constant Mode (CM) waveforms to synthesize spectrum-compatible beampatterns is crucial for Multiple-Input Multiple-Output (MIMO) radar systems. Unlike traditional Fixed Position Antenna (FPA) arrays, we propose a Movable Antenna (MA) enhanced MIMO radar that improves sensing performance by introducing new degrees of freedom (DoFs). This flexibility allows for the adjustment of each antenna’s position based on sensing requirements and optimizing the allocation of sensing power. Specifically, we co-design the transmit waveform and Antenna Position Vector (APV) to minimize the Mean Squared Error (MSE) of beampattern synthesis while adhering to constraints on MA array position, spectrum compatibility, and CM. This leads to a non-convex Quadratic-Constrained Quartic-Programming (QCQP) problem with two highly coupled variables. To solve this problem, we propose an Exact-Penalized-Product-Manifold (EPPM) method. First, we construct a Product Positivity Complex-Circle Manifold (PPC2M) space based on the constraint features of MA array position and CM, projecting both the APV and transmit waveform onto this space. Then, we employ a smoothing technique to transform the spectrum compatibility constraint into an exact penalty function, converting the problem into an unconstrained one on the PPC2M space. Finally, we derive a Parallel Conjugate Gradient (PCG) method to optimize the APV and transmit waveform in parallel. Simulation results show that compared with the existing FPA-based method, the proposed method reduces the beampattern sidelobe by 2.3 dB and the beampatterns matching MSE by about 4 dB.


  • Open Access

    Codesign of Transmit Waveform and Receive Filter with Similarity Constraints for FDA-MIMO Radar

    by Qiping Zhang, Jinfeng Hu, Xin Tai, Yongfeng Zuo, Huiyong Li, Kai Zhong, Chaohai Li

    The codesign of the receive filter and transmit waveform under similarity constraints is one of the key technologies in frequency diverse array multiple-input multiple-output (FDA-MIMO) radar systems. This paper discusses the design of constant modulus waveforms and filters aimed at maximizing the signal-to-interference-and-noise ratio (SINR). The problem’s non-convexity renders it challenging to solve. Existing studies have typically employed relaxation-based methods, which inevitably introduce relaxation errors that degrade system performance. To address these issues, we propose an optimization framework based on the joint complex circle manifold–complex sphere manifold space (JCCM-CSMS). Firstly, the similarity constraint is converted into the penalty term in the objective function using an adaptive penalty strategy. Then, JCCM-CSMS is constructed to satisfy the waveform constant modulus constraint and filter norm constraint. The problem is projected into it and transformed into an unconstrained optimization problem. Finally, the Riemannian limited-memory Broyden–Fletcher–Goldfarb–Shanno (RL-BFGS) algorithm is employed to optimize the variables in parallel. Simulation results demonstrate that our method achieves a 0.6 dB improvement in SINR compared to existing methods while maintaining competitive computational efficiency. Additionally, waveform similarity was also analyzed.


  • Open Access

    Solving a class of multi-scale elliptic PDEs by means of Fourier-based mixed physics informed neural networks

    by Xi'An Li, Jinran Wu, You-Gan Wang, Xin Tai, Jianhua Xu

    Deep neural networks have garnered widespread attention due to their simplicity and flexibility in the fields of engineering and scientific calculation. In this study, we probe into solving a class of elliptic partial differential equations(PDEs) with multiple scales by utilizing Fourier-based mixed physics informed neural networks(dubbed FMPINN), its solver is configured as a multi-scale deep neural network. In contrast to the classical PINN method, a dual (flux) variable about the rough coefficient of PDEs is introduced to avoid the ill-condition of neural tangent kernel matrix caused by the oscillating coefficient of multi-scale PDEs. Therefore, apart from the physical conservation laws, the discrepancy between the auxiliary variables and the gradients of multi-scale coefficients is incorporated into the cost function, then obtaining a satisfactory solution of PDEs by minimizing the defined loss through some optimization methods. Additionally, a trigonometric activation function is introduced for FMPINN, which is suited for representing the derivatives of complex target functions. Handling the input data by Fourier feature mapping will effectively improve the capacity of deep neural networks to solve high-frequency problems. Finally, to validate the efficiency and robustness of the proposed FMPINN algorithm, we present several numerical examples of multi-scale problems in various dimensional Euclidean spaces. These examples cover both low-frequency and high-frequency oscillation cases, demonstrating the effectiveness of our approach. All code and data accompanying this manuscript will be made publicly available at \href{https://github.com/Blue-Giant/FMPINN}{https://github.com/Blue-Giant/FMPINN}.