Zhengxiang Ma

Zhengxiang Ma

Auto-generated profile, based on publicly available information

Zhengxiang Ma

Huawei Technologies (United States)

Research Focus & Metrics

Academic Career Length

30Years (1997-Present)

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

Representative Publications

View All Publications
15 February 2022 in IEEE Transactions on Wireless Communications
Verified
Journal Article
Closed access
WUWolfgang Utschick
Wolfgang Utschick
VRValentina Rizzello
Valentina Rizzello
MJMichael Joham
Michael Joham
ZMZhengxiang Ma
Zhengxiang Ma
LPLeonard Piazzi
Leonard Piazzi
We propose an innovative machine learning-based technique to address the problem of channel acquisition at the base station in frequency division duplex systems. In this context, the base station reconstructs the full channel state information in the downlink frequency range based on limited downlink channel state information feedback from the mobile terminal. The channel state information recovery is based on a convolutional neural network which is trained exclusively on collected channel state samples acquired in the uplink frequency domain. No acquisition of training samples in the downlink frequency range is required at all. Finally, after a detailed presentation and analysis of the proposed technique and its performance, the “reciprocity of distribution” assumption of the convolutional neural network that is central to the proposed approach is validated with an analysis based on the maximum mean discrepancy metric.

1 June 1997 in IEEE Transactions on Applied Superconductivity
Verified
Journal Article
Closed access
ZMZhengxiang Ma
Zhengxiang Ma
Elida de ObaldíaElida de Obaldía
Elida de Obaldía
GHG. Hampel
G. Hampel
PPP. Polakos
P. Polakos
PMP. Mankiewich
P. Mankiewich
JBJ. Beall
J. Beall
In an effort to develop HTS superconducting filters with sufficient power handling capability for PCS (Personal Communication Services) base station transmit applications, we have undertaken a study of the power dependence of large area YBCO thin films on LaAlO/sub 3/ substrates. We employed a coplanar-wave-guide (CPW) resonator technique to obtain the changes of loss and inductance versus circulating microwave currents in the films. Data have been collected on uniform large area (2" diameter) films grown by coevaporation and off-axis sputtering techniques under varying deposition conditions. We found correlations between the RF power dependence and other film properties such as penetration depth and crystal structure. The most intrinsic sample, from the coevaporation technique, characterized by the smallest penetration depth, good orthorhombicity and absence of tetragonal phase, shows the least amount of nonlinearity. Such correlations can be used to prescreen films for fabrication and monitor the film production line. However, films from coevaporation and off-axis sputtering show very distinct power dependent behaviors.

1 January 2015 in Journal of Communications
Verified
Journal Article
Closed access
ZMZhengxiang Ma
Zhengxiang Ma
TCTiejun Chen
Tiejun Chen
Min ZhangMin Zhang
Min Zhang
PKPawel D. Kecerski
Pawel D. Kecerski
SDShuping Dang
Shuping Dang

3 October 2021 in arXiv
Verified
Preprint
Open access
PDF file
WUWolfgang Utschick
Wolfgang Utschick
VRValentina Rizzello
Valentina Rizzello
MJMichael Joham
Michael Joham
ZMZhengxiang Ma
Zhengxiang Ma
LPLeonard Piazzi
Leonard Piazzi
We propose an innovative machine learning-based technique to address the problem of channel acquisition at the base station in frequency division duplex systems. In this context, the base station reconstructs the full channel state information in the downlink frequency range based on limited downlink channel state information feedback from the mobile terminal. The channel state information recovery is based on a convolutional neural network which is trained exclusively on collected channel state samples acquired in the uplink frequency domain. No acquisition of training samples in the downlink frequency range is required at all. Finally, after a detailed presentation and analysis of the proposed technique and its performance, the "transfer learning'' assumption of the convolutional neural network that is central to the proposed approach is validated with an analysis based on the maximum mean discrepancy metric.

18 December 2024
Verified
Preprint
Open access
PDF file
ZMZhengxiang Ma
Zhengxiang Ma
The paper is well-written with a clear explanation of the concepts and procedures that are easy to follow. The Numerical Results section could be further improved if the following issues could be addressed: * There appears to be a significant RMSE...

17 November 2023
Verified
Conference Paper
Closed access
HYHuanyu Yang
Huanyu Yang
AWAihui Wang
Aihui Wang
ZMZhengxiang Ma
Zhengxiang Ma
JYJun Yu
Jun Yu
The giant magnetostrictive actuator (GMA) is an intelligent actuator that utilizes magnetostrictive material as its core, renowned for its exceptional performance, making it extensively employed across various industries. Therefore, the modeling of hysteresis nonlinearity has been the focus of this field. The asymmetric PrandtI-Ishlinskii (PI) model, based on the Hammerstein structure, is employed to represent the nonlinear links within the system. Additionally, for capturing dynamic links in the Hammerstein structure, particularly those associated with high frequency signals, a second-order transfer function is utilized. Determine the model coefficients using the system toolbox. Finally, the developed dynamic rate-dependent hysteresis model is validated by an experimental platform. It is concluded that the asymmetric PI model with Hammerstein structure has higher accuracy, especially for larger frequency excitation signals.