Shuangqing Wei

Prof. Dr. Shuangqing Wei


Research Focus & Metrics

Academic Career Length

21Years (2006-Present)

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

Representative Publications

View All Publications
28 June 2021 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
YJYue Jin
Yue Jin
SWShuangqing Wei
Shuangqing Wei
JYJian Yuan
Jian Yuan
XZXudong Zhang
Xudong Zhang
We solve an important and challenging cooperative navigation control problem, Multiagent Navigation to Unassigned Multiple targets (MNUM) in unknown environments with minimal time and without collision. Conventional methods are based on multiagent path planning that requires building an environment map and expensive real-time path planning computations. In this article, we formulate MNUM as a stochastic game and devise a novel multiagent deep reinforcement learning (MADRL) algorithm to learn an end-to-end solution, which directly maps raw sensor data to control signals. Once learned, the policy can be deployed onto each agent, and thereby, the expensive online planning computations can be offloaded. However, to solve MNUM, traditional MADRL suffers from large policy solution space and nonstationary environment when agents make decisions independently and concurrently. Accordingly, we propose a hierarchical and stable MADRL algorithm. The hierarchical learning part introduces a two-layer policy model to reduce the solution space and uses an interlaced learning paradigm to learn two coupled policies. In the stable learning part, we propose to learn an extended action-value function that implicitly incorporates estimations of other agents' actions, based on which the environment's nonstationarity caused by other agents' changing policies can be alleviated. Extensive experiments demonstrate that our method can converge in a fast way and generate more efficient cooperative navigation policies than comparable methods.

19 October 2020 in IEEE Transactions on Information Forensics and Security
Verified
Journal Article
Closed access
XYXinchun Yu
Xinchun Yu
SWShuangqing Wei
Shuangqing Wei
YLYuan Luo
Yuan Luo
It is well known that finite blocklength analysis plays an important role in evaluating performances of communication systems in practical settings. This paper considers the achievability and converse bounds on the maximal channel coding rate (throughput) at a given blocklength and error probability in covert communication over AWGN channels. The covert constraint is given in terms of an upper bound on total variation distance (TVD) between the distributions of eavesdropped signals at an adversary with and without presence of active and legitimate communication, respectively. For the achievability, Gaussian random coding scheme is adopted for convenience in the analysis of TVD. The classical results of finite blocklength regime are not applicable in this case. By exploiting and extending canonical approaches, we first present new and more general achievability bounds for random coding schemes under maximal or average probability of error requirements. The general bounds are then applied to covert communication in AWGN channels where codewords are generated from Gaussian distribution while meeting the maximal power constraint. We further show an interesting connection between attaining tight achievability and converse bounds and solving two total variation distance based minimax and maxmin problems. The TVD constraint is analyzed under the given random coding scheme, which induces bounds on the transmission power through divergence inequalities. Further comparison is made between the new achievability bounds and existing ones derived under deterministic codebooks. Our thorough analysis thus leads us to a comprehensive characterization of the attainable throughput in covert communication over AWGN channels.

24 August 2021 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
ICIvan E. Carvajal-Roca
Ivan E. Carvajal-Roca
JWJian Wang
Jian Wang
JDJun Du
Jun Du
SWShuangqing Wei
Shuangqing Wei
The current in-vehicle network lacks a robust security communication mechanism between Electronic Control Units (ECUs). Although different proposed frameworks aim to secure the Controller Area Network (CAN) bus communication, few of them consider relevant aspects such as key refreshing and management, ECU maintenance, and vehicle manufacturing process. Despite the fact that a centralized key distribution provides an efficient mechanism, security and safety fully depend on this centralized key generator, which may lead to a single point of failure problem. In this work, we present a semi-centralized dynamic key management framework that provides a decentralized and dynamic key generation during the vehicle's operation. Our framework not only increases security and safety even if the centralized key generator is offline but also reduces the number of keys, the communication payload inside the network, and the complexity of cryptographic material sharing during the vehicle's lifetime. Moreover, we propose a dynamic generation of session keys for different groups of ECU and external devices, which is a requirement for intelligent vehicles and future Internet of Vehicles. In order to evaluate the performance, we implement our framework on three hardware platforms with different computational power and validate its advantage, efficiency, and flexibility by comparing it with other frameworks.

14 November 2022 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
YQYingrong Qin
Yingrong Qin
CGChen Gao
Chen Gao
YWYue Wang
Yue Wang
SWShuangqing Wei
Shuangqing Wei
DJDepeng Jin
Depeng Jin
LZLin Zhang
Lin Zhang
Point-of-Interest (POI) recommendation has drawn a lot of attention in both academia and industry. It utilizes user check-in data, aiming at recommending unvisited POIs to users. To address the data-sparsity problem, geographical information of POIs is often incorporated into recommender systems. However, most of the existing approaches model geographical impact in an implicit way, in which geographical information is encoded as auxiliary vectors for learning unified representations of users and POIs. Following this paradigm, the embedding of POIs can not reflect geographical similarity directly; thus, an explicit modeling approach is needed as geography is of great importance in POI recommendation. To address challenges in disentangling geographical effect, we proposed a disentangled representation learning method named DIG (short for D isentangled embedding of user I nterest and POIs' G eographical information). Aiming at decoupling the geographical factor and the user interest factor thoroughly, we first proposed a geo-constrained negative sampling strategy, which helps to find reliable negative samples for the two factors. Second, a geo-enhanced soft-weighted loss function was proposed to quantify the trade-off between the two factors in loss computation. Extensive experiments have been conducted on two real-world datasets, and results have demonstrated the significant improvement of DIG at $3.92\% \sim 20.32\%$ on recall, and $2.53\%\sim 11.48\%$ on hit ratio, compared with other state-of-the-art approaches.

21 August 2023 in ACM Transactions on Information Systems
Verified
Journal Article
Open access
YQYingrong Qin
Yingrong Qin
CGChen Gao
Chen Gao
SWShuangqing Wei
Shuangqing Wei
YWYue Wang
Yue Wang
DJDepeng Jin
Depeng Jin
YLYong Li
Yong Li
Knowledge graphs (KGs) can help enhance recommendations, especially for the data-sparsity scenarios with limited user-item interaction data. Due to the strong power of representation learning of graph neural networks (GNNs), recent works of KG-based recommendation deploy GNN models to learn from both knowledge graph and user-item bipartite interaction graph. However, these works have not well considered the hierarchical structure of knowledge graph, leading to sub-optimal results. Despite the benefit of hierarchical structure, leveraging it is challenging since the structure is always partly-observed. In this work, we first propose to reveal unknown hierarchical structures with a supervised signal detection method and then exploit the hierarchical structure with disentangling representation learning. We conduct experiments on two large-scale datasets, of which the results well verify the superiority and rationality of the proposed method. Further experiments of ablation study with respect to key model designs have demonstrated the effectiveness and rationality of our proposed model. The code is available at https://github.com/tsinghua-fib-lab/HIKE .

1 January 2013 in Journal of Communications
Verified
Journal Article
Closed access
LQLijun Qian
Lijun Qian
XLXiangfang Li
Xiangfang Li
SWShuangqing Wei
Shuangqing Wei