Zhe Li
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
Zhe Li has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
- Yili Xia
School of Information Science and Engineering, Southeast University, Nanjing, China
11 shared publications - Danilo P. Mandic
Imperial College London, London, United Kingdom
8 shared publications - Wenjiang Pei
School of Information Science and Engineering, Southeast University, Nanjing, China
8 shared publications - Wei Deng
Jiaxing Nanhu University
7 shared publications - Honglei Jin
School of Electronic and Information Engineering, Soochow University, Suzhou, China
4 shared publications - Jingen Ni
Soochow University
3 shared publications
Latest Publications
- 15 October 2025
- Published by Elsevier in Signal Processing
- Vol. 239
Augmented complex-valued total least mean phase filters with application to frequency measurement
- 15 October 2025
- Published by Elsevier in Signal Processing
- Vol. 239
Augmented complex-valued total least mean phase filters with application to frequency measurement
- 10 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 73 (1053587X) , 3795-3811
Time-Averaged ACLMS Algorithm for Improper Cyclostationary Inputs: Performance Analysis and Application to Interference-Limited Systems
In interference-limited communications systems, the use of improper signaling has been demonstrated to improve throughput and fairness, as compared to proper signaling. Considering that signals in such scenarios exhibit both impropriety and cyclostationarity, this paper rigorously studies the optimal adaptive filtering of jointly improper cyclostationary signals. Upon rearranging the cyclic frequencies in the conjugate-linear branch of the FREquency SHift (FRESH) filter, we first propose a widely linear estimation model equivalent to the optimal FRESH filtering. This new structure admits the same form as the widely linear estimator derived with stationary signals, facilitating direct utilization of the augmented complex statistics of the jointly improper cyclostationary input and signal of interest. By minimizing the time-averaged mean-squared error, an adaptive algorithm is obtained for the proposed estimation model, referred to as the time-averaged augmented complex least-mean-squares (TA-ACLMS). We establish a full second-order statistical framework to comprehensively assess the error and the weight error vector of the TA-ACLMS at both transient and steady-state stages, and derive the stability bound on the step-size. The performance of the proposed TA-ACLMS is evaluated through a system identification setting and channel estimation in an interference-limited narrowband power line communication system. Simulation results support the analysis.
- 10 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 73 (1053587X) , 3795-3811
Time-Averaged ACLMS Algorithm for Improper Cyclostationary Inputs: Performance Analysis and Application to Interference-Limited Systems
In interference-limited communications systems, the use of improper signaling has been demonstrated to improve throughput and fairness, as compared to proper signaling. Considering that signals in such scenarios exhibit both impropriety and cyclostationarity, this paper rigorously studies the optimal adaptive filtering of jointly improper cyclostationary signals. Upon rearranging the cyclic frequencies in the conjugate-linear branch of the FREquency SHift (FRESH) filter, we first propose a widely linear estimation model equivalent to the optimal FRESH filtering. This new structure admits the same form as the widely linear estimator derived with stationary signals, facilitating direct utilization of the augmented complex statistics of the jointly improper cyclostationary input and signal of interest. By minimizing the time-averaged mean-squared error, an adaptive algorithm is obtained for the proposed estimation model, referred to as the time-averaged augmented complex least-mean-squares (TA-ACLMS). We establish a full second-order statistical framework to comprehensively assess the error and the weight error vector of the TA-ACLMS at both transient and steady-state stages, and derive the stability bound on the step-size. The performance of the proposed TA-ACLMS is evaluated through a system identification setting and channel estimation in an interference-limited narrowband power line communication system. Simulation results support the analysis.
- 14 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 29 (7) , 1634-1638
RIS-Enhanced Beamwidth-Adjustable Sensing Framework: A Stationary Phase Method-Based Approach
This letter introduces a beamwidth-adjustable sensing beamforming in reconfigurable intelligent surface-assisted communication systems. The proposed approach aims to enhance the detection of potential sensing targets within a broad area centered on the communication user. To achieve this, we develop a channel model that correlates channel gain with spatial offsets and utilize the stationary phase method to derive an optimal numerical solution for adjustable width sensing beamforming. To simplify the process, we propose a low-complexity method that allows for adjustable beamwidth with the ability to achieve zero gain at the beam’s center, making it especially well-suited for integrated sensing and communication applications.
- 14 May 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 29 (7) , 1634-1638
RIS-Enhanced Beamwidth-Adjustable Sensing Framework: A Stationary Phase Method-Based Approach
This letter introduces a beamwidth-adjustable sensing beamforming in reconfigurable intelligent surface-assisted communication systems. The proposed approach aims to enhance the detection of potential sensing targets within a broad area centered on the communication user. To achieve this, we develop a channel model that correlates channel gain with spatial offsets and utilize the stationary phase method to derive an optimal numerical solution for adjustable width sensing beamforming. To simplify the process, we propose a low-complexity method that allows for adjustable beamwidth with the ability to achieve zero gain at the beam’s center, making it especially well-suited for integrated sensing and communication applications.
- 13 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (5) , 7808-7820
Sum-Rate Maximization for Uplink Multi-User NOMA With Improper Gaussian Signaling: A Deep Reinforcement Learning Approach
This paper investigates the joint allocation of power and circularity coefficient for an uplink multi-user nonorthogonal multiple access (NOMA) system employing improper Gaussian signaling (IGS) in the presence of imperfect successive interference cancellation. We propose two novel deep reinforcement learning (DRL) approaches to address the weighted sumrate maximization problem under quality of service constraints in this context. Instead of using widely linear transformation to amalgamate power and circularity coefficient into a unified precoding matrix, the two proposed DRL methods, referred to as interdependent deep deterministic policy gradient (I-DDPG) and collaborative DDPG (C-DDPG), both explicitly incorporate the factor of circularity coefficient in the optimization process, thereby enabling full exploitation of the interference management capabilities offered by IGS in the considered uplink NOMA system. Compared to I-DDPG, C-DDPG facilitates both collaborative training and independent adjustments of both decision variables, leading to enhanced generalization ability. Simulation results validate that the proposed DRL approaches achieve significant sum-rate improvement over conventional model-based optimization techniques, in which C-DDPG pushes individual user rates closer to the optimal NOMA boundary. Furthermore, our findings highlight the fact that the sub-optimality of the model-based optimization technique stems from its inability to fully utilize the circularity coefficient, rather than the power.
- 13 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (5) , 7808-7820
Sum-Rate Maximization for Uplink Multi-User NOMA With Improper Gaussian Signaling: A Deep Reinforcement Learning Approach
This paper investigates the joint allocation of power and circularity coefficient for an uplink multi-user nonorthogonal multiple access (NOMA) system employing improper Gaussian signaling (IGS) in the presence of imperfect successive interference cancellation. We propose two novel deep reinforcement learning (DRL) approaches to address the weighted sumrate maximization problem under quality of service constraints in this context. Instead of using widely linear transformation to amalgamate power and circularity coefficient into a unified precoding matrix, the two proposed DRL methods, referred to as interdependent deep deterministic policy gradient (I-DDPG) and collaborative DDPG (C-DDPG), both explicitly incorporate the factor of circularity coefficient in the optimization process, thereby enabling full exploitation of the interference management capabilities offered by IGS in the considered uplink NOMA system. Compared to I-DDPG, C-DDPG facilitates both collaborative training and independent adjustments of both decision variables, leading to enhanced generalization ability. Simulation results validate that the proposed DRL approaches achieve significant sum-rate improvement over conventional model-based optimization techniques, in which C-DDPG pushes individual user rates closer to the optimal NOMA boundary. Furthermore, our findings highlight the fact that the sub-optimality of the model-based optimization technique stems from its inability to fully utilize the circularity coefficient, rather than the power.
- 7 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 29 (1) , 50-54
Improper Gaussian Signaling for Secrecy Transmission in RIS-Aided Downlink NOMA
This letter investigates the physical layer security of a downlink non-orthogonal multiple access (NOMA) system aided by a reconfigurable intelligent surface (RIS), where artificial noise (AN) is introduced to confuse the eavesdropper. To enhance secrecy, we employ improper Gaussian signaling (IGS) to alleviate the influence of AN on legitimate users. This is formulated as a sum secrecy rate maximization problem, where the reflecting coefficients of RIS and transmit covariance matrices of legitimate users and eavesdroppers are jointly optimized. An alternating optimization framework aided by the minorization-maximization method is adopted to address this non-convex optimization problem. Numerical results show that IGS significantly outperforms proper Gaussian signaling in improving the secrecy performance of RIS-aided NOMA systems, and such a secrecy advantage is more pronounced in RIS-aided NOMA than in traditional NOMA.
- 7 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Communications Letters
- Vol. 29 (1) , 50-54
Improper Gaussian Signaling for Secrecy Transmission in RIS-Aided Downlink NOMA
This letter investigates the physical layer security of a downlink non-orthogonal multiple access (NOMA) system aided by a reconfigurable intelligent surface (RIS), where artificial noise (AN) is introduced to confuse the eavesdropper. To enhance secrecy, we employ improper Gaussian signaling (IGS) to alleviate the influence of AN on legitimate users. This is formulated as a sum secrecy rate maximization problem, where the reflecting coefficients of RIS and transmit covariance matrices of legitimate users and eavesdroppers are jointly optimized. An alternating optimization framework aided by the minorization-maximization method is adopted to address this non-convex optimization problem. Numerical results show that IGS significantly outperforms proper Gaussian signaling in improving the secrecy performance of RIS-aided NOMA systems, and such a secrecy advantage is more pronounced in RIS-aided NOMA than in traditional NOMA.
- 17 October 2024
- Published by Elsevier in Signal Processing
- Vol. 228
An augmented complex-valued gradient-descent total least-squares algorithm for noncircular signals
- 17 October 2024
- Published by Elsevier in Signal Processing
- Vol. 228
An augmented complex-valued gradient-descent total least-squares algorithm for noncircular signals
- Open Access
- 23 February 2024
- Published by Association for Computing Machinery (ACM)
- p. 102-106
Sum-Rate Optimization of Uplink NOMA with Improper Gaussian Signaling and SIC Imperfection
This paper investigates the weighted sum-rate optimization problem in an uplink multi-user non-orthogonal multiple access (NOMA) systems with improper Gaussian signaling (IGS) and imperfect successive interference cancellation. A comprehensive weighted sum-rate maximization problem is formulated, which optimizes user powers while meeting a total power budget and certain quality-of-service demands. An iterative solution is developed to solve this non-convex problem utilizing a widely linear properizer and the weighted minimum mean square error algorithm. Simulation results based on a realistic uplink NOMA model show considerable throughput improvement achieved by adopting IGS.
- 23 February 2024
- Published by Association for Computing Machinery (ACM)
- p. 102-106
Open AccessSum-Rate Optimization of Uplink NOMA with Improper Gaussian Signaling and SIC Imperfection
This paper investigates the weighted sum-rate optimization problem in an uplink multi-user non-orthogonal multiple access (NOMA) systems with improper Gaussian signaling (IGS) and imperfect successive interference cancellation. A comprehensive weighted sum-rate maximization problem is formulated, which optimizes user powers while meeting a total power budget and certain quality-of-service demands. An iterative solution is developed to solve this non-convex problem utilizing a widely linear properizer and the weighted minimum mean square error algorithm. Simulation results based on a realistic uplink NOMA model show considerable throughput improvement achieved by adopting IGS.
- Open Access
Widely Linear Matched Filter: A Lynchpin towards the Interpretability of Complex-valued CNNs
A recent study on the interpretability of real-valued convolutional neural networks (CNNs) {Stankovic_Mandic_2023CNN} has revealed a direct and physically meaningful link with the task of finding features in data through matched filters. However, applying this paradigm to illuminate the interpretability of complex-valued CNNs meets a formidable obstacle: the extension of matched filtering to a general class of noncircular complex-valued data, referred to here as the widely linear matched filter (WLMF), has been only implicit in the literature. To this end, to establish the interpretability of the operation of complex-valued CNNs, we introduce a general WLMF paradigm, provide its solution and undertake analysis of its performance. For rigor, our WLMF solution is derived without imposing any assumption on the probability density of noise. The theoretical advantages of the WLMF over its standard strictly linear counterpart (SLMF) are provided in terms of their output signal-to-noise-ratios (SNRs), with WLMF consistently exhibiting enhanced SNR. Moreover, the lower bound on the SNR gain of WLMF is derived, together with condition to attain this bound. This serves to revisit the convolution-activation-pooling chain in complex-valued CNNs through the lens of matched filtering, which reveals the potential of WLMFs to provide physical interpretability and enhance explainability of general complex-valued CNNs. Simulations demonstrate the agreement between the theoretical and numerical results.
Open AccessWidely Linear Matched Filter: A Lynchpin towards the Interpretability of Complex-valued CNNs
A recent study on the interpretability of real-valued convolutional neural networks (CNNs) {Stankovic_Mandic_2023CNN} has revealed a direct and physically meaningful link with the task of finding features in data through matched filters. However, applying this paradigm to illuminate the interpretability of complex-valued CNNs meets a formidable obstacle: the extension of matched filtering to a general class of noncircular complex-valued data, referred to here as the widely linear matched filter (WLMF), has been only implicit in the literature. To this end, to establish the interpretability of the operation of complex-valued CNNs, we introduce a general WLMF paradigm, provide its solution and undertake analysis of its performance. For rigor, our WLMF solution is derived without imposing any assumption on the probability density of noise. The theoretical advantages of the WLMF over its standard strictly linear counterpart (SLMF) are provided in terms of their output signal-to-noise-ratios (SNRs), with WLMF consistently exhibiting enhanced SNR. Moreover, the lower bound on the SNR gain of WLMF is derived, together with condition to attain this bound. This serves to revisit the convolution-activation-pooling chain in complex-valued CNNs through the lens of matched filtering, which reveals the potential of WLMFs to provide physical interpretability and enhance explainability of general complex-valued CNNs. Simulations demonstrate the agreement between the theoretical and numerical results.