Xiaofeng Ma
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Xiaofeng Ma has a total of 64 co-authors. Here are the top co-authors by number of shared publications:
- Weixing Sheng
School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
19 shared publications - Yubing Han
School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
12 shared publications - Shuang Qiu
School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing, China
5 shared publications - Shuoshuo Song
Shanghai Jiao Tong University
5 shared publications - Renli Zhang
School of Electronic and Optical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China
4 shared publications - Wei-Xing Sheng
Nanjing University of Science and Technology
4 shared publications
Latest Publications
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 74 (11) , 10399-10408
Precise Array Manifold Based on Spherical Harmonic Domain Mutual Coupling Model
- 1 November 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 74 (11) , 10399-10408
Precise Array Manifold Based on Spherical Harmonic Domain Mutual Coupling Model
- 8 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 10708-10719
Wide-FOV Multitarget 2-D DOA Estimation Using Subarray-Partitioned Time-Modulated Planar Arrays
High-resolution two-dimensional (2D) direction-of arrival (DOA) estimation is essential for large-scale sensing and communication systems. However, existing array architectures and estimation methods face significant challenges in simultaneously achieving real-time high-performance DOA estimation with low hardware cost and power consumption, as well as good scalability for large-scale arrays. To address these issues, a subarray partitioned time-modulated planar array (TMPA) architecture is proposed, where time modulation is performed in parallel across subarrays. This design shortens the modulation period, preserves sufficient bandwidth, and enables scalable implementation with low hardware complexity. Based on the proposed architecture, the DOA estimation ambiguity is analyzed, and a two-stage DOA estimation algorithm, termed TMPA-TS-DOA, is developed. Coarse estimation is first conducted in the intra-subarray harmonic beamspace, followed by fine estimation in the inter-subarray convolutional harmonic beamspace, thereby reducing processing dimensionality and computational complexity. Moreover, by exploiting the rotational invariance of the steering vectors in the inter-subarray convolutional harmonic beamspace, a 2D unitary ESPRIT (U-ESPRIT) algorithm is extended to the proposed model for further complexity reduction. Simulation results demonstrate that the proposed method enables reliable wide field-of-view (FOV) multi-target 2D DOA estimation with improved multi-target resolution and lower computational complexity compared with full-element FD-MUSIC and TMA MUSIC, while maintaining high DOA estimation accuracy.
- 8 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 62 (00189251) , 10708-10719
Wide-FOV Multitarget 2-D DOA Estimation Using Subarray-Partitioned Time-Modulated Planar Arrays
High-resolution two-dimensional (2D) direction-of arrival (DOA) estimation is essential for large-scale sensing and communication systems. However, existing array architectures and estimation methods face significant challenges in simultaneously achieving real-time high-performance DOA estimation with low hardware cost and power consumption, as well as good scalability for large-scale arrays. To address these issues, a subarray partitioned time-modulated planar array (TMPA) architecture is proposed, where time modulation is performed in parallel across subarrays. This design shortens the modulation period, preserves sufficient bandwidth, and enables scalable implementation with low hardware complexity. Based on the proposed architecture, the DOA estimation ambiguity is analyzed, and a two-stage DOA estimation algorithm, termed TMPA-TS-DOA, is developed. Coarse estimation is first conducted in the intra-subarray harmonic beamspace, followed by fine estimation in the inter-subarray convolutional harmonic beamspace, thereby reducing processing dimensionality and computational complexity. Moreover, by exploiting the rotational invariance of the steering vectors in the inter-subarray convolutional harmonic beamspace, a 2D unitary ESPRIT (U-ESPRIT) algorithm is extended to the proposed model for further complexity reduction. Simulation results demonstrate that the proposed method enables reliable wide field-of-view (FOV) multi-target 2D DOA estimation with improved multi-target resolution and lower computational complexity compared with full-element FD-MUSIC and TMA MUSIC, while maintaining high DOA estimation accuracy.
- 4 March 2026
- Published by Elsevier in Signal Processing
- Vol. 245
Robust adaptive beamforming of efficient integral reconstruction with piecewise Gauss-Legendre quadrature
- 4 March 2026
- Published by Elsevier in Signal Processing
- Vol. 245
Robust adaptive beamforming of efficient integral reconstruction with piecewise Gauss-Legendre quadrature
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 32 (10709908) , 3969-3973
Revelation and Elimination of Redundant Spectral Searches in MUSIC
In this letter, the spatial-spectral peak search process of the classical multiple signal classification (MUSIC) algorithm is revealed to inherently involve 50% computational redundancy through an analysis of the MUSIC spectrum function, i.e., only half-spatial-domain spectral searches are essential to construct the complete full-spatial-domain spectrum. Thus, the spectral-search framework of MUSIC is refined by the proposed half spectral search MUSIC (HSS-MUSIC), which eliminates the redundant searches and preserves identical outputs of MUSIC. Derivations and simulations demonstrate that compared with MUSIC, the proposed HSS-MUSIC does not exhibit any deterioration in both array structure requirement and DOA estimation performance, but achieves an approximately 50% reduction in computational load.
- 9 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 32 (10709908) , 3969-3973
Revelation and Elimination of Redundant Spectral Searches in MUSIC
In this letter, the spatial-spectral peak search process of the classical multiple signal classification (MUSIC) algorithm is revealed to inherently involve 50% computational redundancy through an analysis of the MUSIC spectrum function, i.e., only half-spatial-domain spectral searches are essential to construct the complete full-spatial-domain spectrum. Thus, the spectral-search framework of MUSIC is refined by the proposed half spectral search MUSIC (HSS-MUSIC), which eliminates the redundant searches and preserves identical outputs of MUSIC. Derivations and simulations demonstrate that compared with MUSIC, the proposed HSS-MUSIC does not exhibit any deterioration in both array structure requirement and DOA estimation performance, but achieves an approximately 50% reduction in computational load.
- 8 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 74 (1) , 1174-1179
Compact Parametric Expression of the True Array Manifold
The accurate parametric array manifold is the theoretical basis of array signal processing technology. In this paper, a compact parametric model of the true array manifold is proposed. A virtual extended array is established by splitting each antenna into a collection of finite-length dipoles, of which the mutual coupling effect is described using a nonlinear coupling model based on Taylor expansion. The resulting composite manifold has a simple form composed of an expanded virtual isolated manifold and a broad coupling matrix, which can provides an extremely accurate representation of the true array manifold. Numerical examples verify that the accuracy of the proposed composite manifold is two or three orders of magnitude better than that of several other state-of-the-art manifolds. The proposed composite manifold can be applied to all kinds of array signal processing algorithms directly to enhance the capabilities of antenna arrays in practical applications by replacing the overly simplified and imprecise manifold.
- 8 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Antennas and Propagation
- Vol. 74 (1) , 1174-1179
Compact Parametric Expression of the True Array Manifold
The accurate parametric array manifold is the theoretical basis of array signal processing technology. In this paper, a compact parametric model of the true array manifold is proposed. A virtual extended array is established by splitting each antenna into a collection of finite-length dipoles, of which the mutual coupling effect is described using a nonlinear coupling model based on Taylor expansion. The resulting composite manifold has a simple form composed of an expanded virtual isolated manifold and a broad coupling matrix, which can provides an extremely accurate representation of the true array manifold. Numerical examples verify that the accuracy of the proposed composite manifold is two or three orders of magnitude better than that of several other state-of-the-art manifolds. The proposed composite manifold can be applied to all kinds of array signal processing algorithms directly to enhance the capabilities of antenna arrays in practical applications by replacing the overly simplified and imprecise manifold.
- 12 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 61 (6) , 18219-18233
Unambiguous Manifold Analysis of Spatially Spread Dual-Polarized Dipole Array
In this article, the comprehensive analysis of the spatially spread dual-polarized dipole array (SSDDA) manifold ambiguity is conducted, such that the sufficient conditions for the unambiguous array manifold geomrtries of the linear SSDDA (LSSDDA) and planar SSDDA (PSSDDA) are derived. Consequently, if one of two linear horizontal dipole configurations is involved in the LSSDDA, i.e., 1) the configuraion owning adjacent horizontal dipoles, 2) the configuration owing co-prime horizontal dipoles' spacings, the unambiguous 1-D directionof-arrival (DOA) and polarization estimations can be achieved apart from vertically polarized signals. Similarly, if one of two linear vertical dipole configurations is involved in the LSSDDA, i.e., 1) the configuraion owning adjacent vertical dipoles, 2) the configuration owing co-prime vertical dipoles' spacings, the unambiguous 1-D DOA and polarization estimations can be achieved apart from horizontally polarized signals. Furthermore, it is also confirmed that when one of two planar horizontal dipole configurations is integrated in the PSSDDA, i.e., 1) the configuration owning adjacent horizontal dipoles along x- and y-axes, 2) the configuration owning co-prime horizontal dipoles' spacings along x- and y-axes, the unambiguous 2-D DOA and polarization estimations are available except for the condition pph(θ, φ, γ, η) = 0 and ppv(θ, φ, γ, η), 0. Meanwhile, when one of two planar vertical dipole configurations is integrated in the PSSDDA, i.e., 1) the configuration owning adjacent vertical dipoles along x- and y-axes, 2) the configuration owning co-prime vertical dipoles' spacings along x- and y-axes, the unambiguous 2-D DOA and polarization estimations are available except for pph(θ, φ, γ, η), 0 and ppv(θ, φ, γ, η) = 0. Numerical simulations are performed to verify the proof of unambiguous LSSDDA and PSSDDA manifold geometries.
- 12 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Aerospace and Electronic Systems
- Vol. 61 (6) , 18219-18233
Unambiguous Manifold Analysis of Spatially Spread Dual-Polarized Dipole Array
In this article, the comprehensive analysis of the spatially spread dual-polarized dipole array (SSDDA) manifold ambiguity is conducted, such that the sufficient conditions for the unambiguous array manifold geomrtries of the linear SSDDA (LSSDDA) and planar SSDDA (PSSDDA) are derived. Consequently, if one of two linear horizontal dipole configurations is involved in the LSSDDA, i.e., 1) the configuraion owning adjacent horizontal dipoles, 2) the configuration owing co-prime horizontal dipoles' spacings, the unambiguous 1-D directionof-arrival (DOA) and polarization estimations can be achieved apart from vertically polarized signals. Similarly, if one of two linear vertical dipole configurations is involved in the LSSDDA, i.e., 1) the configuraion owning adjacent vertical dipoles, 2) the configuration owing co-prime vertical dipoles' spacings, the unambiguous 1-D DOA and polarization estimations can be achieved apart from horizontally polarized signals. Furthermore, it is also confirmed that when one of two planar horizontal dipole configurations is integrated in the PSSDDA, i.e., 1) the configuration owning adjacent horizontal dipoles along x- and y-axes, 2) the configuration owning co-prime horizontal dipoles' spacings along x- and y-axes, the unambiguous 2-D DOA and polarization estimations are available except for the condition pph(θ, φ, γ, η) = 0 and ppv(θ, φ, γ, η), 0. Meanwhile, when one of two planar vertical dipole configurations is integrated in the PSSDDA, i.e., 1) the configuration owning adjacent vertical dipoles along x- and y-axes, 2) the configuration owning co-prime vertical dipoles' spacings along x- and y-axes, the unambiguous 2-D DOA and polarization estimations are available except for pph(θ, φ, γ, η), 0 and ppv(θ, φ, γ, η) = 0. Numerical simulations are performed to verify the proof of unambiguous LSSDDA and PSSDDA manifold geometries.
- 7 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. PP (10709908) , 1-5
ESPRIT for Time-Modulated Array with Unidirectional Phase Center Motion
The direction of arrival (DOA) estimation algorithms for time-modulated arrays (TMAs) with extremely low hardware complexity are attractive. However, due to the time modulation operation performed on each antenna of the TMA to generate the harmonic beams, the rotational invariance is invalid in the harmonic domain. As a result, the estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm cannot be directly applied to the TMA. In this letter, an ESPRIT algorithm for the TMA (TMA-ESPRIT) is proposed for the first time. First, the unidirectional phase center motion (UPCM) modulation is employed and the harmonic domain signal subspace is constructed accordingly. Then, a derived harmonic coefficient matrix (HCM) is performed on the harmonic domain signal subspace to reconstruct the rotational invariance. Numerical simulations demonstrate the superior performance of the TMA-ESPRIT algorithm in terms of the computational complexity and the multi-target resolution.
- 7 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. PP (10709908) , 1-5
ESPRIT for Time-Modulated Array with Unidirectional Phase Center Motion
The direction of arrival (DOA) estimation algorithms for time-modulated arrays (TMAs) with extremely low hardware complexity are attractive. However, due to the time modulation operation performed on each antenna of the TMA to generate the harmonic beams, the rotational invariance is invalid in the harmonic domain. As a result, the estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm cannot be directly applied to the TMA. In this letter, an ESPRIT algorithm for the TMA (TMA-ESPRIT) is proposed for the first time. First, the unidirectional phase center motion (UPCM) modulation is employed and the harmonic domain signal subspace is constructed accordingly. Then, a derived harmonic coefficient matrix (HCM) is performed on the harmonic domain signal subspace to reconstruct the rotational invariance. Numerical simulations demonstrate the superior performance of the TMA-ESPRIT algorithm in terms of the computational complexity and the multi-target resolution.
- 1 July 2025
- Published by Elsevier in Signal Processing
- Vol. 238
Joint DOA, polarization and mutual coupling parameters estimation based on linear crossed-dipole array
- 1 July 2025
- Published by Elsevier in Signal Processing
- Vol. 238
Joint DOA, polarization and mutual coupling parameters estimation based on linear crossed-dipole array
- 6 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Antennas and Wireless Propagation Letters
- Vol. 24 (9) , 2904-2908
Accurate Parametric Analytic Manifold Using a Nonlinear Mutual Coupling Model
The precise mutual coupling model is the key to obtain an accurate parametric analytic manifold, which is the prerequisite to achieve super-resolution direction of arrival estimation and highprecision beamforming. The widely used mutual coupling model is linear and cannot describe the direction-dependent characteristics of mutual coupling, which leads to serious performance deterioration of array signal processing algorithms. In this letter, a new non-linear mutual coupling model named high-order broad mutual coupling matrix (MCM) is established based on the Taylor expansion to describe the directional dependent coupling relationship between two adjacent antenna elements. Then, an accurate parametric analytic manifold using a finite-order approximation of the high-order broad MCM is yield by the antenna pattern reconstruction method. Numerical examples verify that the array manifold reconstruction errors can be reduced exponentially by the proposed parametric analytic manifold. Especially when the piecewise fitting strategy is adopted, the reconstruction errors of the fourth-order approximated analytic manifold can be reduced to less than 1% of those of the conventional array manifold.
- 6 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Antennas and Wireless Propagation Letters
- Vol. 24 (9) , 2904-2908
Accurate Parametric Analytic Manifold Using a Nonlinear Mutual Coupling Model
The precise mutual coupling model is the key to obtain an accurate parametric analytic manifold, which is the prerequisite to achieve super-resolution direction of arrival estimation and highprecision beamforming. The widely used mutual coupling model is linear and cannot describe the direction-dependent characteristics of mutual coupling, which leads to serious performance deterioration of array signal processing algorithms. In this letter, a new non-linear mutual coupling model named high-order broad mutual coupling matrix (MCM) is established based on the Taylor expansion to describe the directional dependent coupling relationship between two adjacent antenna elements. Then, an accurate parametric analytic manifold using a finite-order approximation of the high-order broad MCM is yield by the antenna pattern reconstruction method. Numerical examples verify that the array manifold reconstruction errors can be reduced exponentially by the proposed parametric analytic manifold. Especially when the piecewise fitting strategy is adopted, the reconstruction errors of the fourth-order approximated analytic manifold can be reduced to less than 1% of those of the conventional array manifold.