Xin He
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
Research Subjects
Xin He has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Jiangzhou Wang
Southeast University
5 shared publications - Lei Huang
College of Electronic and Information Engineering, Shenzhen University, Shenzhen, China
5 shared publications - Yi Gong
Southern University of Science and Technology
4 shared publications Lei Huang
3 shared publicationsJiangzhou Wang
2 shared publicationsYik-Chung Wu
Department of Electrical and Electronic Engineering, The University of Hong Kong, Hong Kong
2 shared publications
Latest Publications
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15 (21622337) , 2904-2908
Stacked Intelligent Metasurface-Assisted Holographic MIMO Communications via Unfolded Algorithm
Stacked intelligent metasurfaces (SIMs) can enhance transmission efficiency in holographic MIMO (HMIMO) systems, but their multi-layer structure poses computational challenges. This paper proposes an unfolded gradient projection (GP) method, where the step sizes are trained by neural networks, leading to faster convergence for the SIM-assisted system. Simulation results demonstrate that the unfolded GP method achieves significant computational complexity reduction and faster convergence compared to conventional GP and alternating optimization benchmarks. Furthermore, the method exhibits robustness against channel estimation errors and strong scalability.
- 29 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 15 (21622337) , 2904-2908
Stacked Intelligent Metasurface-Assisted Holographic MIMO Communications via Unfolded Algorithm
Stacked intelligent metasurfaces (SIMs) can enhance transmission efficiency in holographic MIMO (HMIMO) systems, but their multi-layer structure poses computational challenges. This paper proposes an unfolded gradient projection (GP) method, where the step sizes are trained by neural networks, leading to faster convergence for the SIM-assisted system. Simulation results demonstrate that the unfolded GP method achieves significant computational complexity reduction and faster convergence compared to conventional GP and alternating optimization benchmarks. Furthermore, the method exhibits robustness against channel estimation errors and strong scalability.
- 6 August 2025
- Published by Elsevier in Digital Signal Processing
- Vol. 168
Dual-function MIMO communication and radar with multi-target priorities and fairness
- 6 August 2025
- Published by Elsevier in Digital Signal Processing
- Vol. 168
Dual-function MIMO communication and radar with multi-target priorities and fairness
- 9 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 14 (7) , 1914-1918
Reconfigurable Holographic Surface Beamforming via Unrolled Algorithm
Satellite communication network may plays a important role in the future sixth-generation (6G) wireless communication. Reconfigurable Holographic Surface (RHS) is a promising solution for satellite communication network. However, the RHS induces the problem of high computational complexity. To mitigate this problem, an unrolled gradient projection (GP) method is proposed in this letter. The unrolled GP method uses neural networks to train the step size in the traditional GP method, which significantly speed up the convergence rate. Extensive simulations show that the unrolled GP method runs faster and achieves a higher sum rate than the fractional programming (FP) and phased array methods.
- 9 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Wireless Communications Letters
- Vol. 14 (7) , 1914-1918
Reconfigurable Holographic Surface Beamforming via Unrolled Algorithm
Satellite communication network may plays a important role in the future sixth-generation (6G) wireless communication. Reconfigurable Holographic Surface (RHS) is a promising solution for satellite communication network. However, the RHS induces the problem of high computational complexity. To mitigate this problem, an unrolled gradient projection (GP) method is proposed in this letter. The unrolled GP method uses neural networks to train the step size in the traditional GP method, which significantly speed up the convergence rate. Extensive simulations show that the unrolled GP method runs faster and achieves a higher sum rate than the fractional programming (FP) and phased array methods.
- 1 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (2) , 3463-3467
Linear Complexity Holographic Beamforming for Satellite Broadcasting
To compensate for the severe pathloss in satellite broadcasting system, the phased array with a half wavelength aperture is the conventional scheme. However, it is difficult to further scale up the phased array to provide high data rate communication. The reconfigurable holographic surface (RHS) with continuous or quasi-continuous aperture is a promising alternative to phased array for satellite communications. Unlike the amplitude ratio controlled RHS, the amplitude-controlled RHS is firstly modelled in the satellite broadcasting system. To reduce the computational complexity of the RHS sum rate maximization problem, the unrolled gradient projection (GP) method is proposed. The simulation results show that the RHS scheme has larger sum rate than the phased array method, and the unrolled GP method for RHS is faster than the phased array method.
- 1 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (2) , 3463-3467
Linear Complexity Holographic Beamforming for Satellite Broadcasting
To compensate for the severe pathloss in satellite broadcasting system, the phased array with a half wavelength aperture is the conventional scheme. However, it is difficult to further scale up the phased array to provide high data rate communication. The reconfigurable holographic surface (RHS) with continuous or quasi-continuous aperture is a promising alternative to phased array for satellite communications. Unlike the amplitude ratio controlled RHS, the amplitude-controlled RHS is firstly modelled in the satellite broadcasting system. To reduce the computational complexity of the RHS sum rate maximization problem, the unrolled gradient projection (GP) method is proposed. The simulation results show that the RHS scheme has larger sum rate than the phased array method, and the unrolled GP method for RHS is faster than the phased array method.
- Open Access
- 1 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 91177-91186
Integrating Energy Harvesting, Anti-Eavesdropping, and Communication in a Hybrid Beamforming System
This paper proposes a scheme to integrate data communication, wireless energy harvesting and anti-eavesdropping functions into a hybrid beamforming downlink system to satisfy the demand for expanded functionalities in a communication system. To solve this difficult nonconvex problem, the Semidefinite Relaxation (SDR) method is employed to obtain the digital transmitter, the SDR method and penalty convex-concave procedure (CCP) method are employed to obtain the analog transmitter. Simulation results shows four conclusions about the system. Firstly, the larger the SINR target or the energy harvest target, the more transmit power is required. Secondly, the lower the data rate obtained by the eavesdropper, the more transmit power is required. Thirdly, the SDR method and the penalty CCP method complement each other. Finally, the transmit power can be reduced by increasing the number of RF chains or transmit antennas.
- 1 July 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 91177-91186
Open AccessIntegrating Energy Harvesting, Anti-Eavesdropping, and Communication in a Hybrid Beamforming System
This paper proposes a scheme to integrate data communication, wireless energy harvesting and anti-eavesdropping functions into a hybrid beamforming downlink system to satisfy the demand for expanded functionalities in a communication system. To solve this difficult nonconvex problem, the Semidefinite Relaxation (SDR) method is employed to obtain the digital transmitter, the SDR method and penalty convex-concave procedure (CCP) method are employed to obtain the analog transmitter. Simulation results shows four conclusions about the system. Firstly, the larger the SINR target or the energy harvest target, the more transmit power is required. Secondly, the lower the data rate obtained by the eavesdropper, the more transmit power is required. Thirdly, the SDR method and the penalty CCP method complement each other. Finally, the transmit power can be reduced by increasing the number of RF chains or transmit antennas.
- 21 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 73 (7) , 10783-10787
Learn to Optimize RIS Aided Hybrid Beamforming With Out-of-Distribution Generalization
Owing to the manually fixed step size, the conventional gradient projection (GP) method requires relatively long time to solve the reconfigurable intelligent surface (RIS) aided hybrid beamforming problem. In order to speed up the GP method, we propose to learn the step sizes by using deep learning. Since the proposed deep learning architecture has a coordinate ascent structure, every step in the deep learning is explainable. Due to the simple multi-layer architecture, the proposed unrolled GP method has a strong out-of-distribution generalization capability. Under a single training setting, the unrolled GP approach is tested under thirty nine different out-of-distribution settings. The extensive simulation results show that the unrolled GP method has larger achievable rate than the GP method under middle-to-high signal-to-noise ratio (SNR) settings, and the proposed method is ten times faster than the GP method for all settings. Codes are available url https://github.com/hexingit/RIS-aided-HBF.
- 21 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 73 (7) , 10783-10787
Learn to Optimize RIS Aided Hybrid Beamforming With Out-of-Distribution Generalization
Owing to the manually fixed step size, the conventional gradient projection (GP) method requires relatively long time to solve the reconfigurable intelligent surface (RIS) aided hybrid beamforming problem. In order to speed up the GP method, we propose to learn the step sizes by using deep learning. Since the proposed deep learning architecture has a coordinate ascent structure, every step in the deep learning is explainable. Due to the simple multi-layer architecture, the proposed unrolled GP method has a strong out-of-distribution generalization capability. Under a single training setting, the unrolled GP approach is tested under thirty nine different out-of-distribution settings. The extensive simulation results show that the unrolled GP method has larger achievable rate than the GP method under middle-to-high signal-to-noise ratio (SNR) settings, and the proposed method is ten times faster than the GP method for all settings. Codes are available url https://github.com/hexingit/RIS-aided-HBF.
- 17 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23 (3) , 1742-1754
Efficient Algorithms for RIS Aided Hybrid Beamforming With MSE Constraints
In this paper, to stabilize the users’ quality of service (QoS), the symbol detection mean squared error (MSE) constrained hybrid analog and digital beamforming is proposed in millimeter wave (mmWave) system, and the reconfigurable intelligent surface (RIS) is proposed to assist the mmWave system. The inner majorization-minimization (iMM) method is proposed to obtain analog transmitter, RIS and analog receivers, and the alternating direction method of multipliers (ADMM) method is proposed to obtain digital transmitter. The proposed iMM and ADMM methods are faster than the semidefinite relaxation (SDR) and interior point methods, respectively. In order to counter against the changing large-scale path loss, the iMM method is helped by channel normalization to reduce the computational complexity, while the ADMM method is helped by the adaptive parameterizations to robust against the changing large-scale path loss. Simulation results show that the computational times of the proposed method are much faster than other methods, the proposed method is robust against the changing large-scale path loss, and the user’s bit error rate (BER) is stable under small to medium channel estimation errors.
- 17 July 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Wireless Communications
- Vol. 23 (3) , 1742-1754
Efficient Algorithms for RIS Aided Hybrid Beamforming With MSE Constraints
In this paper, to stabilize the users’ quality of service (QoS), the symbol detection mean squared error (MSE) constrained hybrid analog and digital beamforming is proposed in millimeter wave (mmWave) system, and the reconfigurable intelligent surface (RIS) is proposed to assist the mmWave system. The inner majorization-minimization (iMM) method is proposed to obtain analog transmitter, RIS and analog receivers, and the alternating direction method of multipliers (ADMM) method is proposed to obtain digital transmitter. The proposed iMM and ADMM methods are faster than the semidefinite relaxation (SDR) and interior point methods, respectively. In order to counter against the changing large-scale path loss, the iMM method is helped by channel normalization to reduce the computational complexity, while the ADMM method is helped by the adaptive parameterizations to robust against the changing large-scale path loss. Simulation results show that the computational times of the proposed method are much faster than other methods, the proposed method is robust against the changing large-scale path loss, and the user’s bit error rate (BER) is stable under small to medium channel estimation errors.
- 29 March 2023
- Published by Elsevier in Digital Signal Processing
- Vol. 137
MIMO radar design with joint constant-modulus and any p-norm similarity constraints
- 29 March 2023
- Published by Elsevier in Digital Signal Processing
- Vol. 137
MIMO radar design with joint constant-modulus and any p-norm similarity constraints
- 27 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 70 (1053587X) , 5237-5250
QCQP With Extra Constant Modulus Constraints: Theory and Application to SINR Constrained Mmwave Hybrid Beamforming
The constant modulus constraint is widely used in analog beamforming, hybrid beamforming, intelligent reflecting surface design, and radar waveform design. The quadratically constrained quadratic programming (QCQP) problem is also widely used in signal processing. However, the QCQP with extra constant modulus constraints was not systematically studied in mathematic programming and signal processing. For example, the multiple quality of service (QoS) constrained analog beamforming is rare, while the QoS constrained digital beamforming methods are abundant. We propose to tackle the QCQP with extra constant modulus constraints problem by solving a series of subproblems with linear programming (LP) under extra constant modulus constraints. Under mild condition, the strong duality between the LP with extra constant modulus constraints and its dual problem is established. Then, by using the optimal solutions from the subproblems, the QCQP with extra constant modulus constraints problem is solved with a monotonically converged algorithm, and all converged solutions are Karush-Kuhn-Tucker (KKT) points. As an application of the proposed method, the signal-to-interference-plus-noise-ratio (SINR) constrained hybrid beamforming is solved. Simulation results show that the transmit power of the proposed method is similar to that of the semidefinite relaxation (SDR) method, while the computational time of proposed method is much smaller than that of the SDR method. The proposed method always provides feasible solution, while the SDR method does not always provide feasible solution.
- 27 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 70 (1053587X) , 5237-5250
QCQP With Extra Constant Modulus Constraints: Theory and Application to SINR Constrained Mmwave Hybrid Beamforming
The constant modulus constraint is widely used in analog beamforming, hybrid beamforming, intelligent reflecting surface design, and radar waveform design. The quadratically constrained quadratic programming (QCQP) problem is also widely used in signal processing. However, the QCQP with extra constant modulus constraints was not systematically studied in mathematic programming and signal processing. For example, the multiple quality of service (QoS) constrained analog beamforming is rare, while the QoS constrained digital beamforming methods are abundant. We propose to tackle the QCQP with extra constant modulus constraints problem by solving a series of subproblems with linear programming (LP) under extra constant modulus constraints. Under mild condition, the strong duality between the LP with extra constant modulus constraints and its dual problem is established. Then, by using the optimal solutions from the subproblems, the QCQP with extra constant modulus constraints problem is solved with a monotonically converged algorithm, and all converged solutions are Karush-Kuhn-Tucker (KKT) points. As an application of the proposed method, the signal-to-interference-plus-noise-ratio (SINR) constrained hybrid beamforming is solved. Simulation results show that the transmit power of the proposed method is similar to that of the semidefinite relaxation (SDR) method, while the computational time of proposed method is much smaller than that of the SDR method. The proposed method always provides feasible solution, while the SDR method does not always provide feasible solution.
- Open Access
- 8 April 2022
- Published by Institution of Engineering and Technology (IET) in IET Signal Processing
- Vol. 16 (7) , 838-850
Transceiver coexistence design of MIMO radar and MIMO communication under Gaussian model uncertainty
This paper proposes a transceiver coexistence design of multiple‐input multiple‐output (MIMO) radar and MIMO communication under Gaussian model uncertainty. To tackle the mutual interference and the model uncertainty problems, the transceivers of the radar system and the communication system are jointly designed to minimise the total transmit power of the radar system and the communication system, while the signal‐to‐interference‐plus‐noise‐ratios (SINRs) or effective SINR (ESINR) of the radar system and the mean square error (MSE) of the communication system are guaranteed with small outage probability. Safe approximations are proposed for the probabilistic constraints, and a converged iterative algorithm is proposed for the joint transceiver design. Owing to the MSE formulation, the convex relaxation for the communication transmitter is proved to be tight. Furthermore, the tight communication relaxation provides an effective way to check whether the radar error is majorly due to numeric error or the possible rank‐one relaxation error. Extensive simulation results show that the proposed iterative algorithm converges, the possible radar rank‐one relaxation error is trivial with respect to the numeric error, and the quality of service in terms of SINR, ESINR, MSE and bit error rate (BER) in the proposed coexistence system are robust against model uncertainty with the cost of using extra transmit power.
- 8 April 2022
- Published by Institution of Engineering and Technology (IET) in IET Signal Processing
- Vol. 16 (7) , 838-850
Open AccessTransceiver coexistence design of MIMO radar and MIMO communication under Gaussian model uncertainty
This paper proposes a transceiver coexistence design of multiple‐input multiple‐output (MIMO) radar and MIMO communication under Gaussian model uncertainty. To tackle the mutual interference and the model uncertainty problems, the transceivers of the radar system and the communication system are jointly designed to minimise the total transmit power of the radar system and the communication system, while the signal‐to‐interference‐plus‐noise‐ratios (SINRs) or effective SINR (ESINR) of the radar system and the mean square error (MSE) of the communication system are guaranteed with small outage probability. Safe approximations are proposed for the probabilistic constraints, and a converged iterative algorithm is proposed for the joint transceiver design. Owing to the MSE formulation, the convex relaxation for the communication transmitter is proved to be tight. Furthermore, the tight communication relaxation provides an effective way to check whether the radar error is majorly due to numeric error or the possible rank‐one relaxation error. Extensive simulation results show that the proposed iterative algorithm converges, the possible radar rank‐one relaxation error is trivial with respect to the numeric error, and the quality of service in terms of SINR, ESINR, MSE and bit error rate (BER) in the proposed coexistence system are robust against model uncertainty with the cost of using extra transmit power.