Wenqiang Pu

Dr. Wenqiang Pu

Shenzhen Research Institute of Big Data

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

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

Representative Publications

View All Publications
6 July 2020 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
JYJunkun Yan
Junkun Yan
WPWenqiang Pu
Wenqiang Pu
Shenghua ZhouShenghua Zhou
Shenghua Zhou
Hongwei LiuHongwei Liu
Hongwei Liu
MGMaria Sabrina Greco
Maria Sabrina Greco
In this paper, two optimal resource allocation schemes are developed for asynchronous multiple targets tracking (MTT) in heterogeneous radar network. The key idea of the heterogeneous resource allocation (HRA) schemes is to coordinate the heterogeneous transmit resource (transmit power, dwell time, etc.) of different types of radars for better resource utilization efficiency. We use the Bayesian Cramr-Rao lower bound (BCRLB) as a metric function to quantify the target tracking performance, and build the following two HRA schemes: For given system resource budget, (1) Minimize the total resource consumption for given BCRLB requirements on multiple targets; (2) Maximize the overall MTT accuracy. Instead of updating the state of each target recursively at different measurement arrival times, we combine multiple asynchronous measurements into a single composite measurement, and use it as an input of the tracking filter for state estimation. In such a case, target tracking BCRLB no longer needs to be recursively calculated, and thus we can formulate the HRA schemes as two convex optimization problems. We subsequently design two efficient methods to solve these problems by exploring their unique structures. Simulation results demonstrate that the HRA processes can either provide smaller overall MTT BCRLB for given resource budgets, or require less resource to establish the same tracking performance for multiple targets.

6 April 2021 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
JYJunkun Yan
Junkun Yan
JDJinhui Dai
Jinhui Dai
WPWenqiang Pu
Wenqiang Pu
Hongwei LiuHongwei Liu
Hongwei Liu
MGMaria Greco
Maria Greco
In this paper, a target capacity based resource optimization (TC-RO) scheme is developed for multiple target tracking (MTT) application in radar networks. The key idea of this scheme is to coordinate the transmit power and dell time resource usage of multiple radars in order to increase the number of the targets that can be tracked with predetermined accuracy requirements. We adopt the Bayesian Cramér-Rao lower bound as a metric function to quantify the MTT accuracies, and build the TC-RO scheme as a non-smooth and non-convex optimization problem. To deal with this problem, we design an efficient three-step solution technique which incorporates relaxation and fine-tuning process. Specifically, we first relax the resulting optimization problem as a smooth one by applying sigmoid-type transformation to its objective, and then develop an appropriate method to find a local minimum to the relaxed non-convex problem with guaranteed convergence. After that, the local minimum of the relaxed problem is used as an initial point and a fine-tuning process is performed to search for a reasonable feasible solution to the original non-smooth optimization problem. Simulation results demonstrate that the proposed TC-RO scheme can greatly increase the target capacity of the radar network when compared with the traditional uniform allocation scheme.

29 May 2018 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
JYJunkun Yan
Junkun Yan
WPWenqiang Pu
Wenqiang Pu
Hongwei LiuHongwei Liu
Hongwei Liu
BJBo JIU
Bo JIU
ZBZheng Bao
Zheng Bao
Taking into account the probabilistic uncertainty on the target radar cross section (RCS) parameter, a robust chance constrained power allocation (RCC-PA) scheme is presented for multiple target localization in colocated multiple-input multiple-output radar system. Such a system adopts a multibeam working mode, in which multiple simultaneous transmit beams are synthesized to illuminate multiple targets independently. We formulate the RCC-PA problem into a chance constrained programming model, where the total power consumption of the multiple beams is minimized to meet a specified multi-target localization accuracy requirement with high probability. Various target RCS fluctuation models have been discussed with different forms of probability distributions. We analytically show that the chance constrained programming problem for each RCS fluctuation model can equivalently be formulated as a deterministic convex optimization problem. Then, by formulating the Karush-Kuhn-Tuckers conditions, we transform the convex optimization problem into a nonlinear equation solving problem, and then solve it by using the bisection method. Simulation results show that our RCC-PA scheme can enhance the power utilization efficiency, and is more robust than the existing deterministic PA schemes.

14 January 2022 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
HSHaoran Sun
Haoran Sun
WPWenqiang Pu
Wenqiang Pu
XFXiao Fu
Xiao Fu
TCTsung-Hui Chang
Tsung-Hui Chang
MHMingyi Hong
Mingyi Hong
There has been a growing interest in developing data-driven, and in particular deep neural network (DNN) based methods for modern communication tasks. These methods achieve state-of-the-art performance for a few popular wireless resource allocation problems, while requiring less computational efforts, less resources for acquiring channel state information (CSI), etc. However, it is often challenging for these approaches to learn in a dynamic environment. This work develops a new approach that enables data-driven methods to continuously learn and optimize wireless resource allocation in a dynamic environment. Specifically, we consider an “episodically dynamic” setting where the environment statistics change in “episodes,” and in each episode the environment is stationary. We propose to build the notion of continual learning (CL) into wireless system design, so that the learning model can incrementally adapt to the new episodes, without forgetting knowledge learned from the previous episodes. We demonstrate the effectiveness of the CL approach by integrating it with three popular DNN based models for power control, beamforming and multi-user MIMO, respectively, and testing using both synthetic and ray-tracing based data sets. These numerical results show that the proposed CL approach is not only able to adapt to the new scenarios quickly and seamlessly, but importantly, it also maintains high performance over the previously encountered scenarios as well.

16 May 2022 in IEEE Transactions on Aerospace and Electronic Systems
Verified
Journal Article
Closed access
KLKang Li
Kang Li
BJBo JIU
Bo JIU
WPWenqiang Pu
Wenqiang Pu
Hongwei LiuHongwei Liu
Hongwei Liu
XPXiaojun Peng
Xiaojun Peng
One emerging issue in modern electronic warfare is the competition between the radar and jammer, which in principle can be viewed as a non-cooperative game with two players. In practice, the interaction between the radar and jammer involves multiple rounds as well as partial observation. This makes the competition become a dynamic game with imperfect information. Anti-jamming strategy design for such kind of game is still unclear. In this work, the competition between a frequency agile (FA) radar and a transmit/receive time-sharing jammer is considered. We utilize an extensive-form game with imperfect information to model the multiple rounds interaction between the radar and jammer. For the established extensive-form game, finding Nash Equilibrium (NE) strategies is a computationally intractable task since the number of information states grows exponentially with respect to game stages. Instead, a sampled based learning method called Neural Fictitious Self Play (NFSP) algorithm is used to find approximate NE strategies. Simulation results show that approximate NE strategies can be obtained and outperform the common elementary and advanced strategies from the perspective of detection performance.

23 January 2019 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
JYJunkun Yan
Junkun Yan
WPWenqiang Pu
Wenqiang Pu
JDJinhui Dai
Jinhui Dai
Hongwei LiuHongwei Liu
Hongwei Liu
ZBZheng Bao
Zheng Bao
To facilitate the ability of phased array radar to manage both its search and track (SAT) tasks simultaneously within a predetermined illumination time budget, a resource allocation (RA) scheme for integrated SAT application is built in this paper. We formulate the RA scheme as a bi-objective constrained optimization framework, and use Pareto's theory to determine its multiple Pareto optimal solutions. With these Pareto solutions, one can find a suitable tradeoff between SAT tasks, and correspondingly choose an illumination scheme, for an arbitrary application demand. By exploiting the unique structure of the bi-objective optimization problem, we strictly prove that the multiple Pareto solutions with cardinality can be obtained by parallelly solving convex minimax problems. These problems are shown to correlate with the search and track tasks independently, and are correspondingly solved by the well-known linear programming methods and a proposed minimax solving algorithm, respectively. Finally, some numerical results are provided to illustrate the effectiveness and reveal the intrinsic mechanism of the Pareto theory based bi-objective RA strategy.