Abstract: Automatic modulation classification (AMC) serves a challenging yet crucial role in wireless communications. Despite deep learning-based approaches being widely used in signal processing, they are challenged by signal distribution variations, especially in various channel conditions. In this paper, we introduce an adversarial transfer framework named frequency-learning adversarial networks (FLANs) based on transfer learning for cross-scenario signal classification. This method uses the stability in the frequency spectrum by introducing a frequency adaptation (FA) technique to incorporate target channel information into source-domain signals. To address the unpredictable interference in the channel, a fitting channel adaptation (FCA) module is used to reduce the difference between the source and target domains caused by variations in the channel environment. Experimental results illustrate that FLANs outperforms state-of-the-art transfer approaches, demonstrating an improved top-1 classification accuracy by about 5.2 percentage points in high signal-to-noise ratio (SNR) scenes on a cross-scenario real collected dataset CSRC2023.

基于迁移学习下跨场景信号调制分类的频谱学习对抗网络

[1]Alarabi A, Alkishriwo OAS, 2021. Modulation classification based on statistical features and artificial neural network. Proc IEEE 1^st Int Maghreb Meeting of the Conf on Sciences and Techniques of Automatic Control and Computer Engineering MI-STA, p.748-751.

[2]Bu K, He Y, Jing XJ, et al., 2020. Adversarial transfer learning for deep learning based automatic modulation classification. IEEE Signal Process Lett, 27:880-884.

[3]Chen ZZ, Cui H, Xiang JY, et al., 2022. SigNet: a novel deep learning framework for radio signal classification. IEEE Trans Cogn Commun Netw, 8(2):529-541.

[4]G N, Vijayan V, Jose R, 2021. Performance analysis of modulation classification using machine learning. Proc 8^th Int Conf on Smart Computing and Communications, p.70-74.

[5]Ganin Y, Ustinova E, Ajakan H, et al., 2015. Domain-adversarial training of neural networks. https://arxiv.org/abs/1505.07818

[6]Goodfellow IJ, Pouget-Abadie J, Mirza M, et al., 2014. Generative adversarial nets. Proc 27^th Int Conf on Neural Information Processing Systems, p.2672-2680.

[7]He KM, Zhang XY, Ren SQ, et al., 2016. Deep residual learning for image recognition. Proc IEEE Conf on Computer Vision and Pattern Recognition, p.770-778.

[8]Huang S, Dai R, Huang JJ, et al., 2020. Automatic modulation classification using gated recurrent residual network. IEEE Int Things J, 7(8):7795-7807.

[9]Lin WS, Hou DB, Huang JS, et al., 2023. Transfer learning for automatic modulation recognition using a few modulated signal samples. IEEE Trans Veh Technol, 72(9):12391-12395.

[10]Lin Y, Zhao HJ, Ma XF, et al., 2021. Adversarial attacks in modulation recognition with convolutional neural networks. IEEE Trans Reliab, 70(1):389-401.

[11]Long MS, Cao Y, Wang JM, et al., 2015. Learning transferable features with deep adaptation networks. Proc 32^nd Int Conf on Machine Learning, p.97-105.

[12]Lukito WD, Rashad FE, Hamid EY, 2021. Multi features-based baseband modulation classification using support vector machine. Proc Int Conf on Radar, Antenna, Microwave, Electronics, and Telecommunications, p.227-231.

[13]Meng F, Chen P, Wu LN, et al., 2018. Automatic modulation classification: a deep learning enabled approach. IEEE Trans Veh Technol, 67(11):10760-10772.

[14]Nam H, Lee H, Park J, et al., 2021. Reducing domain gap by reducing style bias. Proc IEEE/CVF Conf on Computer Vision and Pattern Recognition, p.8686-8695.

[15]O’Shea TJ, Corgan J, Clancy TC, 2016. Convolutional radio modulation recognition networks. Proc 17^th Int Conf on Engineering Applications of Neural Networks, p.213-226.

[16]O’Shea TJ, Roy T, Clancy TC, 2018. Over-the-air deep learning based radio signal classification. IEEE J Sel Top Signal Process, 12(1):168-179.

[17]Pan SJ, Yang Q, 2010. A survey on transfer learning. IEEE Trans Knowl Data Eng, 22(10):1345-1359.

[18]Perenda E, Rajendran S, Bovet G, et al., 2021. Learning the unknown: improving modulation classification performance in unseen scenarios. Proc IEEE Conf on Computer Communications, p.1-10.

[19]Rajendran S, Meert W, Giustiniano D, et al., 2018. Deep learning models for wireless signal classification with distributed low-cost spectrum sensors. IEEE Trans Cogn Commun Netw, 4(3):433-445.

[20]Ramjee S, Ju ST, Yang DY, et al., 2021. Ensemble wrapper subsampling for deep modulation classification. IEEE Trans Cogn Commun Netw, 7(4):1156-1170.

[21]Salama AA, Morsy ME, Darwish SH, et al., 2022. A novel SVM-based automatic modulation classifier. Proc Int Telecommunications Conf, p.1-3.

[22]Tan X, Xie ZD, Yuan XW, et al., 2022. Small sample signal modulation recognition based on higher-order cumulants and CatBoost. Proc 7^th Int Conf on Communication, Image and Signal Processing, p.324-329.

[23]Tang B, Tu Y, Zhang ZY, et al., 2018. Digital signal modulation classification with data augmentation using generative adversarial nets in cognitive radio networks. IEEE Access, 6:15713-15722.

[24]Tzeng E, Hoffman J, Saenko K, et al., 2017. Adversarial discriminative domain adaptation. Proc IEEE Conf on Computer Vision and Pattern Recognition, p.2962-2971.

[25]Wang MY, Lin Y, Tian Q, et al., 2021. Transfer learning promotes 6G wireless communications: recent advances and future challenges. IEEE Trans Reliab, 70(2):790-807.

[26]Wang Q, Du PF, Yang JY, et al., 2019. Transferred deep learning based waveform recognition for cognitive passive radar. Signal Process, 155:259-267.

[27]Wang XC, Lv Z, Gao X, et al., 2022. Pervasive wireless channel modeling theory and applications to 6G GBSMs for all frequency bands and all scenarios. IEEE Trans Veh Technol, 71(9):9159-9173.

[28]Wang ZG, Yan WZ, Oates T, 2017. Time series classification from scratch with deep neural networks: a strong baseline. Proc Int Joint Conf on Neural Networks, p.1578-1585.

[29]Xiao WX, Ding ZM, Liu HF, 2021. Implicit semantic response alignment for partial domain adaptation. Proc 35^th Int Conf on Neural Information Processing Systems, p.1059.

[30]Xu Y, Li DZ, Wang ZY, et al., 2019. A deep learning method based on convolutional neural network for automatic modulation classification of wireless signals. Wirel Netw, 25(7):3735-3746.

[31]Xu ZW, Han GJ, Liu L, et al., 2023. A lightweight specific emitter identification model for IIoT devices based on adaptive broad learning. IEEE Trans Ind Inform, 19(5):7066-7075.

[32]Yao ZS, Fu X, Guo LT, et al., 2023. Few-shot specific emitter identification using asymmetric masked auto-encoder. IEEE Commun Lett, 27(10):2657-2661.

[33]Zhao ZQ, Khan FN, Li YB, et al., 2022. Application and comparison of active and transfer learning approaches for modulation format classification in visible light communication systems. Opt Expr, 30(10):16351-16361.

[34]Zheng Y, Yu L, Yang RR, et al., 2021. A general 3D non-stationary massive MIMO GBSM for 6G communication systems. Proc IEEE Wireless Communications and Networking Conf, p.1-6.

[35]Zhou HJ, Wang X, Bai J, et al., 2022. Modulation signal recognition based on selective knowledge transfer. Proc IEEE Global Communications Conf, p.1875-1880.

[36]Zhu YC, Zhuang FZ, Wang JD, et al., 2021. Deep subdomain adaptation network for image classification. IEEE Trans Neur Netw Learn Syst, 32(4):1713-1722.