Can supervise: YES
Wu, K, Ni, W, Zhang, J, Liu, RP & Guo, YJ 2020, 'Refinement of Optimal Interpolation Factor for DFT Interpolated Frequency Estimator', IEEE COMMUNICATIONS LETTERS, vol. 24, no. 4, pp. 782-786.View/Download from: Publisher's site
Wu, K, Ni, W, Zhang, JA, Liu, RP & Guo, YJ 2020, 'Secrecy Rate Analysis for Millimeter-Wave Lens Antenna Array Transmission', IEEE COMMUNICATIONS LETTERS, vol. 24, no. 2, pp. 272-276.View/Download from: Publisher's site
Wu, K, Ni, W, Su, T, Liu, RP & Guo, Y 2019, 'Exploiting Spatial-Wideband Effect for Fast AoA Estimation at Lens Antenna Array', IEEE Journal on Selected Topics in Signal Processing, vol. 13, no. 5, pp. 902-917.View/Download from: Publisher's site
© 2007-2012 IEEE. Energy-efficient, highly integrated lens antenna arrays (LAAs) have found widespread applications in wideband millimeter wave or terahertz communications, localization and tracking, and wireless power transfer. Accurate estimation of angle-of-arrival (AoA) is key to those applications, but has been hindered by a spatial-wideband effect in wideband systems. This paper proposes to exploit (rather than circumventing) the spatial-wideband effect to develop a fast and accurate AoA estimation approach for LAAs. Specifically, we unveil new spatial-frequency patterns based on the spatial-wideband effect, and establish one-to-one mappings between the patterns and the strongest discrete Fourier transform (DFT) beam containing the AoA. With the strongest DFT beam identified, we propose a method to estimate the AoA uniquely and accurately, using only a few training symbols. This is achieved by deriving a new one-to-one mapping between the AoA and the set of DFT beams judiciously selected based on the strongest. In the case that an impinging path is uniformly distributed in [0,2π], simulations show that the proposed algorithm is able to reduce the mean squared error of the AoA estimation by as much as 82.1% while reducing the number of required symbols by 93.2$, as compared to existing techniques. The algorithm can also increase the spectral efficiency by 89% when the average SNR is 20 dB at each antenna of the receiver.
Wu, K, Ni, W, Su, T, Liu, RP & Guo, YJ 2019, 'Efficient angle-of-arrival estimation of lens antenna arrays for wireless information and power transfer', IEEE Journal on Selected Areas in Communications, vol. 37, no. 1, pp. 116-130.View/Download from: Publisher's site
© 1983-2012 IEEE. Antenna design and angle-of-arrival (AoA) estimation are critical to the efficiency of wireless information and power transfer. The AoA estimation is challenging for energy-efficient lens antenna arrays (LAAs), due to discrete sets of fixed discrete Fourier transform (DFT) beams. This paper presents a novel fast and accurate approach for the AoA estimation of LAAs. The key idea is that we prove the two differential outputs of three adjacent lens beams, referred to as 'DFT beam differences (DBDs),' that are the strongest at the two sides of an AoA. They are easy to identify and robust to noises, and their powers are proved to provide an accurate estimate of the AoA. Another important aspect is a new beam synthesis technique which produces different beam widths based on DFT beams and practical 1-bit phase shifts in real time. As a result, the angular region containing the AoA can exponentially narrow down, and the two strongest DBDs can be quickly identified. The proposed approach can operate in coupling with successive interference cancellation to estimate the AoAs of multiple paths. Simulations show that the proposed approach is able to outperform the state of the art by orders of magnitude in terms of accuracy. The power transfer efficiency can be dramatically improved.
Wu, K, Ni, W, Su, T, Liu, RP & Guo, YJ 2019, 'Expeditious estimation of angle-of-arrival for hybrid butler matrix arrays', IEEE Transactions on Wireless Communications, vol. 18, no. 4, pp. 2170-2185.View/Download from: Publisher's site
© 2002-2012 IEEE. Arrays of Butler matrices provide a promising front-end design for massive MIMO transceivers with low cost and low complexity. However, this advanced design does not necessarily translate to effective applications, unless the angle-of-arrival (AoA) of signals avails to the Butler matrices. This paper presents an efficient approach to the unprecedented AoA estimation for the arrays of Butler matrices. Specifically, we design a new beam synthesis method to recursively narrow down and increasingly focus on the angular region of interest, and hence achieving robust estimation of the phase offset between Butler matrices. With the phase offset canceled in the received signals, we are able to identify the set of critical Butler beams with the dominating effect on the AoA estimation, and estimate the AoA accordingly with minimum signaling. The mean squared error of the proposed estimation is analyzed in the presence of non-negligible noises, with closed-form lower bounds derived. Validated by simulations, the proposed algorithm is able to indistinguishably approach the lower bounds, and significantly outperforms the state-of-the-art developed for discrete antenna arrays by orders of magnitude in terms of accuracy, especially in low signal-to-noise regimes.
Wu, K, Ni, W, Su, T, Liu, RP & Guo, YJ 2019, 'Recent Breakthroughs on Angle-of-Arrival Estimation for Millimeter-Wave High-Speed Railway Communication', IEEE Communications Magazine, vol. 57, no. 9, pp. 57-63.View/Download from: Publisher's site
© 2019 IEEE. With significantly improved efficiency, largescale hybrid antenna arrays with tens to hundreds of antennas have great potential to support millimeter-wave (mmWave) communication for high-speed railway (HSR) applications. The significant beamforming gains rely on fast and accurate estimation of the angle-of-arrival (AoA), but this can be impeded by the high train speed, the cost/energy oriented design of arrays, and the severe attenuation of mmWave signals. This article reviews these challenges, and discusses the limitations of existing AoA estimation techniques under hybrid antenna array settings. The article further reveals a few recent theoretical breakthroughs that can potentially enable fast and reliable estimation, even based on severely attenuated signals. Under a speed setting of 500 km/h, a performance study is carried out to confirm the significant improvements of estimation accuracy and subsequent beamforming gains as the results of the breakthroughs.
Li, Q, Su, T & Wu, K 2019, 'Accurate DOA Estimation for Large-Scale Uniform Circular Array Using a Single Snapshot', IEEE COMMUNICATIONS LETTERS, vol. 23, no. 2, pp. 302-305.View/Download from: Publisher's site
Wu, K, Ni, W, Su, T, Liu, RP & Guo, YJ 2018, 'Fast and Accurate Estimation of Angle-of-Arrival for Satellite-Borne Wideband Communication System', IEEE Journal on Selected Areas in Communications, vol. 36, no. 2, pp. 314-326.View/Download from: Publisher's site
© 1983-2012 IEEE. Accurate estimation of angle-of-arrival (AoA) is critical to wideband satellite communications, but is susceptible to receive noises and can be ambiguous due to space/cost-effective hybrid antenna array designs with localized analog phased subarrays. As a matter of fact, there has yet to be an unambiguous estimator even for narrow-band systems. This paper proposes a new design of subarray-specific time-varying phase shifts, which enables unambiguous and noise-tolerant estimation of AoA in localized hybrid arrays. Particularly, the new phase shifts deliver deterministic phase changes in the cross-correlations of receive signals between subarrays, and enable the cross-correlations to be coherently accumulated across subarrays and sub-carriers to eliminate ambiguities and tolerate noises. Another important contribution of the paper is that we optimize the frequency interval for coherent accumulation across sub-carriers, leveraging between estimation errors, and accumulation gains. Evident from simulations, our approach is able to dramatically improve the estimation accuracy by orders of magnitudes with significantly reduced requirements of complexities and training symbols, as compared with the state of the art. The approach is robust against noises, with estimation errors asymptotically achieving a rigorously developed lower bound.
Wu, K, Ni, W, Su, T, Liu, RP & Guo, YJ 2018, 'Robust Unambiguous Estimation of Angle-of-Arrival in Hybrid Array With Localized Analog Subarrays', IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, vol. 17, no. 5, pp. 2987-3002.View/Download from: Publisher's site
Lv, Q, Su, T & He, X 2018, 'An ISAR Imaging Algorithm for Nonuniformly Rotating Targets With Low SNR Based on Modified Bilinear Parameter Estimation of Cubic Phase Signal', IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS, vol. 54, no. 6, pp. 3108-3124.View/Download from: Publisher's site
Wu, K, Su, T & Li, Q 2017, 'Phase-only Beamforming via Iterative Majorization Minimization', Dianzi Yu Xinxi Xuebao/Journal of Electronics and Information Technology, vol. 39, no. 3, pp. 662-669.View/Download from: Publisher's site
© 2017, Science Press. All right reserved. To improve the interference and noise suppression ability of Phase-Only Beamforming (POB) and the real-time performance to obtain the optimal weight of the uon-convex POB, a phase-only Minimum Variance Distortion-less Response Beamformer (MVDRB) via the Iterative Majorization Minimization (IMM) algorithm is designed. The distortion-less response constraint is transformed into maximizing array response in the desired direction. The upper bound function of the objective function in quadratic form is derived which generates the IMM-POB model. The optimal weight in closed-form for each iteration is further deduced. Simulation analysis illustrates that the proposed IMM-POB can obtain better signal to interference and noise ratio improvement compared with the existing iterative POB algorithms and the optimal MVDRB; its interference suppression ability approximates the optimal MVDRB with better noise suppression ability; it is much less sensitive to the number of array taps compared with the optimal MVDRB; it has strong real-time performance and can be applied to large-scale array.
Lü, Q, Su, T, Wu, K & Zhang, J 2016, 'ISAR imaging algorithm for targets with the complex motion based on the high-order ambiguity function-Lv's distribution', Xi'an Dianzi Keji Daxue Xuebao/Journal of Xidian University, vol. 43, no. 5, pp. 24-30.View/Download from: Publisher's site
© 2016, The Editorial Board of Journal of Xidian University. All right reserved. In ISAR imaging for targets with the complex motion, since the azimuth high order phase terms caused by the time-varying rotation vector will deteriorate the azimuth focusing quality, the traditional RD algorithm and LFM model are not appropriate. Thereby, in the case when the received signal can be modeled as cubic phase signals (CPSs), this paper proposes an ISAR imaging algorithm based on HAF-LVD (high-order ambiguity function-Lv's distribution). First, this algorithm defines a novel double lag parametric instantaneous autocorrelation function, and then applies the scaling operator to remove the coupling and utilizes FFT to achieve the energy accumulation. Finally, the non-searching estimation of the moving parameter and the ISAR images for targets are accomplished by the obtained HAF-LVD. Because of the introduction of the scaling factor, this algorithm can flexibly deal with more changeful and hostile ISAR environment without loss of the anti-noise performance and computational efficiency. Simulation results validate the effectiveness of the ISAR imaging approach.
Wu, K & Su, T 2016, 'Design of generalised sidelobe canceller-based adaptive wideband beamformer without pre-steering delays', Electronics Letters, vol. 52, no. 3, pp. 177-179.View/Download from: Publisher's site
© The Institution of Engineering and Technology 2016. Implementation errors of pre-steering (PS) delays lead to severe performance degradation in the generalised sidelobe canceller-based adaptive wideband beamformer (G-AWBF). Therefore, a G-AWBF without PS delaysisproposed. Toeliminate PS delays, the G-AWBF is designed in the linear space spanned by the generalised steering vector (GSV) which is defined as the Kronecker product of spatial and temporal steering vectors. Based on GSVs, the quiescent weight and blocking matrix are constructed to substitute PS delays. Compared with the conventional G-AWBF, due to the removal of PS delays, an improved signal to interference and noise ratio can beachieved with reduced system complexity. With less adaptive weights and various versions of unconstrained adaptive algorithms being applicable, the proposed G-AWBF outperforms the existing Frost-based AWBF without PS delays in terms of computational load and robustness. Numerical experiments illustrate the effectiveness and superiority of the proposed G-AWBF.
Wu, K & Su, T 2016, 'Design of generalised sidelobe canceller-based adaptive wideband beamformer without pre-steering delays', ELECTRONICS LETTERS, vol. 52, no. 3, pp. 177-178.View/Download from: Publisher's site
Wu, K, Su, T, Jin, B & Yang, T 2016, 'Design of adaptive broadband beamformer with fixed feedback weights', Jilin Daxue Xuebao (Gongxueban)/Journal of Jilin University (Engineering and Technology Edition), vol. 46, no. 1, pp. 303-310.View/Download from: Publisher's site
© 2016, Editorial Board of Jilin University. All right reserved. By introducing a feedback branch with fixed weights, a new adaptive broadband beamformer based on the Generalized Sidelobe Canceller (GSC) is designed. The feedback weights are obtained off-line through approximating the band-pass filter whose passband covers that of the interference. Also, the relationship between the interference suppression capability of the beamformer and the number of poles of the feedback branch is investigated. Compared with existing beamformer, since the feedforward branch of the new beamformer is exactly the same as the GSC, the complexity and amount of calculations of the updating procedure for feedforward weights are reduced. Meanwhile, the fixed feedback weights guarantee its stability. Besides, the flexibility of changing the number of poles strengths its interference suppression capability.
Wu, K & Su, T 2015, 'Design and application of the allpass fractional delay filter', Xi'an Dianzi Keji Daxue Xuebao/Journal of Xidian University, vol. 42, no. 4, pp. 8-13.View/Download from: Publisher's site
©, 2015, Xi'an Dianzi Keji Daxue Xuebao/Journal of Xidian University. All right reserved. By analyzing the design and solving of the allpass fractional delay filter (APFDF) based on the weighted least square method, a new APFDF design procedure adopting the modified genetic algorithm based on elite strategy is proposed, which assures the convergence and stability in obtaining the filter's coefficients. The hybrid genetic operators speed up the convergence of genetic operation towards the global optimal solution and prevent it from pondering around the solution. Simulation results show that the allpass fractional delay filters designed are of low order and high precision. Applied in wideband digital beamforming (WDBF), the aperture transition and frequency dispersion problems are solved successfully and the beam formed approximates the ideal beam.
Wu, K, Su, T, Li, Q & He, XH 2015, 'Design of robust constant beamwidth beamformer with maximal sparsity', Tongxin Xuebao/Journal on Communications, vol. 36, no. 9, pp. 160-168.View/Download from: Publisher's site
©, 2015, Editorial Board of Journal on Communications. All right reserved. To reduce the complexity of broadband array systems, an optimization model was built based on the analysis of the sparsity of the broadband array. The objective function was the convex combination of sensor and TDL sparsity with the constraint of constant beamwidth. By re-weighting the objective function, the sparsity of the array weights increased and converged to the maximal value. The equivalence between the norm constraint on array weights and the sparsity of TDL was proved. Simulation results demonstrate that to obtain the same performance, the number of sensors and TDL of the proposed beamformer needed is decreased which reduces the system complexity, showing that the beamformer designed has great practical value.
© 2016 IEEE. In the traditional range alignment algorithm of inverse synthetic aperture radar (ISAR) imaging, strong data coupling reduces the efficiency of the algorithm extremely. To decouple the data to the largest extent, a fine-grained optimized parallel range alignment algorithm based on logarithmic steps is proposed in this paper, and the parallel implement of ISAR imaging on Graphic Processing Unit (GPU) is designed as well, which provides a new way for real-time ISAR imaging. Experiments were carried out on a blade server with the GPU accelerator (NVIDIA K80) based on simulated target model. Results show that the GPU-based parallel algorithm is faster (up to 5 times) than the original serial algorithm, and the time consuming almost keep constant under increasing sample rate or longer accumulation time, which make it suitable for real-time imaging.
For inverse synthetic aperture radar (ISAR) imaging of targets with complex motions, the received signals have to be modeled as multi-component cubic phase signals (CPSs) after motion compensation. The innovation of this paper includes the introduction of the nonuniform fast Fourier transform (NUFFT) algorithm to obtain highquality ISAR images rapidly. Because of the application of the NUFFT and the SCFT via chirp-z transform (CZT) with respect to the lag-time axis and time axis, respectively, the proposed algorithm can achieve a compromise between performance and low computational complexity in estimating the parameters. Both the theoretical analysis and the simulation examples demonstrate the effectiveness of the proposed algorithm.
Wu, K, Su, T & Pang, J 2014, 'A NEW BROADBAND BEAMFORMER WITH FIXED FEEDBACK WEIGHTS', 2014 12TH INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING (ICSP), 12th IEEE International Conference on Signal Processing (ICSP), IEEE, HangZhou, PEOPLES R CHINA, pp. 271-276.