Zhi Liang

Zhi Liang

Auto-generated profile, based on publicly available information

Zhi Liang

School of Electronics and Information, Northwestern Polytechnical University, Xian, China

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
29 June 2021 in IEEE Communications Letters
Verified
Journal Article
Closed access
ZLZhi Liang
Zhi Liang
Mingliang TaoMingliang Tao
Mingliang Tao
LWLing Wang
Ling Wang
JSJia Su
Jia Su
XYXin Yang
Xin Yang
Automatic modulation recognition (AMR) plays an important role in modern wireless communication. In this letter, a novel framework for AMR is proposed. The ResNeXt network serves as the backbone, and four proposed adaptive attention mechanism modules are incorporated. The time-frequency representations of the received signals are utilized as the inputs of the proposed deep learning (DL) network, and a transfer learning strategy is adopted based on the pre-trained ResNeXt weakly supervised learning (WSL) model. The comparisons with several state-of-the-art techniques on the RadioML2016.10B and RadioML2018.01A datasets show that the proposed framework converges quickly and can achieve higher robustness and 2% to 8% higher accuracy than other state-of-the-art techniques.

31 May 2022 in IEEE Transactions on Cognitive Communications and Networking
Verified
Journal Article
Closed access
ZLZhi Liang
Zhi Liang
Mingliang TaoMingliang Tao
Mingliang Tao
JXJian Xie
Jian Xie
XYXin Yang
Xin Yang
LWLing Wang
Ling Wang
Automatic modulation recognition (AMR) of radio signals is becoming increasingly important due to its key role in wireless communication system management, monitoring, and control. In this paper, we propose an end-to-end AMR framework based on deep learning (DL), named CCNN-Atten. First, the complex-valued convolutional neural network (CCNN) extracts the features of the radio signal, and the feature calibration (FC) module selectively enhances the important features and suppresses irrelevant features. Then, a temporal context capture (TCC) module uses a modified multi-head attention mechanism (MHA) to capture the temporal dependence in the extracted features. The improved MHA mechanism, as a kind of self-attention mechanism, deploys causal convolutions to encode the temporal information of the input features and captures their local temporal relationship. In addition, due to the limited hardware resources in the real scenarios, we also considered a good compromise between recognition accuracy and computational complexity. Experiments were performed with the RadioML2016.10B, RadioML2016.10A, and RadioML2018.01A datasets, demonstrating the ability of the proposed CCNN-Atten to learn more robust features than other state-of-the-art (SOTA) techniques with 1 ~ 10% higher accuracy and a lower computational complexity than the SOTA models. The experimental results also show that CCNN-Atten achieves outstanding performance in dealing with radio signals with a lower sampling rate and small signal observation window.

25 July 2022 in IEEE Communications Letters
Verified
Journal Article
Closed access
ZLZhi Liang
Zhi Liang
JXJian Xie
Jian Xie
XYXin Yang
Xin Yang
Mingliang TaoMingliang Tao
Mingliang Tao
LWLing Wang
Ling Wang
Deep learning has made great progress in the field of radio signal recognition. However, existing works usually assume that testing and training data own the same feature domain, which may not be the case in real-world scenarios. The existing cross-domain radio signal recognition methods based on domain adaptation only consider the global feature alignment. A novel unsupervised adversarial domain-adaptive learning framework named DASig is proposed to address the cross-domain radio signal recognition issue for the data acquired in different channel conditions. To align the fine-grained class distributions of the target and source domains, we designed an efficient combined loss function and iterative self-training strategy via target domain pseudo-labels. Moreover, a non-shared domain self-attention mechanism is designed to preserve domain-specific features. Extensive experiment comparisons with five state-of-the-art (SOTA) methods on both the modulation datasets and protocol datasets show the superiority of the DASig.