Zhongju Wang

Zhongju Wang

Auto-generated profile, based on publicly available information

Zhongju Wang

Hong Kong University of Science and Technology, Hong Kong

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 April 2017 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
ZWZhongju Wang
Zhongju Wang
PBPrabhu Babu
Prabhu Babu
DPDaniel P. Palomar
Daniel P. Palomar
Phase noise correction is crucial to exploit full advantage of orthogonal frequency-division multiplexing (OFDM) in modern high-data-rate communications. OFDM channel estimation with simultaneous phase noise compensation has therefore drawn much attention and stimulated continuing efforts. Existing methods, however, either have not taken into account the fundamental properties of phase noise or are only able to provide estimates of limited applicability owing to considerable computational complexity. In this paper, we have reformulated the joint phase noise and channel estimation problem in the time domain as opposed to existing frequency-domain approaches, which enables us to develop much more efficient algorithms using the majorization-minimization technique. In addition, we propose two methods based on dimensionality reduction and regularization, respectively, that can adapt to various phase noise levels and signal-to-noise ratio and achieve much lower estimation errors than the benchmarks without incurring much additional computational cost. Several numerical examples with phase noise generated by free-running oscillators or phase-locked loops demonstrate that our proposed algorithms outperform existing methods with respect to both computational efficiency and mean squared error within a large range of SNRs.

8 September 2016 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
ZWZhongju Wang
Zhongju Wang
PBPrabhu Babu
Prabhu Babu
DPDaniel P. Palomar
Daniel P. Palomar
Communication systems and radars widely employ sequences of low peak-to-average power ratio (PAR) or unimodulus to meet the hardware constraints and maximize the power efficiency. Numerous works have probed the unimodular sequence design especially attempting to obtain good correlation properties. Regarding channel estimation, however, sequences of such properties do not necessarily qualify for the mission. And tailored unimodular sequences for the specific criterion concerned are more desirable when we have access to the prior knowledge of the channel impulse response. In this paper, we formulate the problem of unimodular sequence design by optimizing minimum mean square error and conditional mutual information, respectively. The problems turn out nonconvex and we develop efficient algorithms based on the majorization-minimization framework with convergence guaranteed. More general, we also examine optimal sequence design with low PAR constraints. Numerical examples demonstrate the improved results of mean square error, signal-to-noise ratio, and conditional mutual information by using our proposed training sequences, with the efficiency of the derived algorithms illustrated.


1 June 2018
Verified
Conference Paper
Closed access
ZZZiping Zhao
Ziping Zhao
RZRui Zhou
Rui Zhou
ZWZhongju Wang
Zhongju Wang
DPDaniel P. Palomar
Daniel P. Palomar
In this paper, the optimal mean-reverting portfolio (MRP) design problem is considered, which plays an important role for the statistical arbitrage (a.k.a. pairs trading) strategy in financial markets. The target of the optimal MRP design is to construct a portfolio from the underlying assets that can exhibit a satisfactory mean reversion property and a desirable variance property. A general problem formulation is proposed by considering these two targets and an investment leverage constraint. To solve this problem, a successive convex approximation method is used. The performance of the proposed model and algorithms are verified by numerical simulations.

1 July 2017
Verified
Conference Paper
Closed access
ZWZhongju Wang
Zhongju Wang
PBPrabhu Babu
Prabhu Babu
DPDaniel P. Palomar
Daniel P. Palomar
Phase noise compensation plays an indispensable role in realizing the potential of orthogonal frequency division multiplexing (OFDM) for high-data-rate communications. Due to various heuristics employed to deal with the particular properties of phase noise, the existing approaches are still performance-limited in terms of accuracy and computational cost. In this paper, we focus on channel estimation stage and formulate the problem of phase noise compensation in the time domain. An efficient algorithm is devised using the majorization-minimization technique with guaranteed convergence. Numerical results with phase noise generated as a Wiener process demonstrate that our proposed algorithm is able to provide substantially improved phase noise and channel estimates with lower mean squared error especially for moderate signal-to-noise ratio (SNR). The computational complexity is also much lower than that of the benchmark.

1 August 2021
Verified
Conference Paper
Closed access
MZMeng Zhang
Meng Zhang
ZWZhongju Wang
Zhongju Wang
Long WangLong Wang
Long Wang
Chao HuangChao Huang
Chao Huang
Xiong LuoXiong Luo
Xiong Luo
This paper proposes an automated lung cancer classification model based on the dielectric and conductivity properties of lung tissues. In the proposed model, Extreme Gradient Boosting (XGBoost) is utilized to perform the classification and Whale Optimization algorithm (WOA) is employed to tune the hyper-parameters. Furthermore, to reduce the negative impact of data imbalance on model training, a multinomial weighted focal (MWF) loss is developed and synthetic minority over-sampling technique (SMOTE) is employed to balance distribution of sub-categories. To validate feasibility of the proposed model, permittivity and conductivity data of 119 lung cancer surgical patients are collected from Nanfang Hospital, China. A total of 229 samples covering five cancer states are used to assess the performance of the proposed model. Meanwhile, SVM and AdaBoost are benchmarked to demonstrate the superiority of the proposed model. Experimental results prove that the proposed model obtains the best classification accuracy and macro-precision compared with other methods. Therefore, the proposed method is effective for lung cancer classification tasks.