{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T01:05:40Z","timestamp":1760144740941,"version":"build-2065373602"},"reference-count":34,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2024,5,8]],"date-time":"2024-05-08T00:00:00Z","timestamp":1715126400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"UniSQ-DSTG Postgraduate Research Scholarship 2021\u20132024","award":["10254"],"award-info":[{"award-number":["10254"]}]},{"name":"Department of Defence, Commonwealth of Australia under DSP Scholarship","award":["10254"],"award-info":[{"award-number":["10254"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>In trainable wireless communications systems, the use of deep learning for over-the-air training aims to address the discontinuity in backpropagation learning caused by the channel environment. The primary methods supporting this learning procedure either directly approximate the backpropagation gradients using techniques derived from reinforcement learning, or explicitly model the channel environment by training a generative channel model. In both cases, over-the-air training of transmitter and receiver requires a feedback channel to sound the channel environment and obtain measurements of the learning objective. The use of continuous feedback not only demands extra system resources but also makes the training process more susceptible to adversarial attacks. Conversely, opting for a feedback-free approach to train the models over the forward link, exclusively on the receiver side, could pose challenges to reliably end the training process without intermittent testing over the actual channel environment. In this article, we propose a novel method for the over-the-air training of wireless communication systems that does not require a feedback channel to train the transmitter and receiver. Random samples are transmitted through the channel environment to train a mixture density network to approximate the channel distribution on the receiver side of the network. The transmitter and receiver models are trained with the resulting channel model, and the transmitter can be deployed after training. We show that the block error rate measurements obtained with the simulated channel are suitable for monitoring as a stopping criterion during the training process. The resulting method is demonstrated to have equivalent performance to the end-to-end autoencoder training on small message sequences.<\/jats:p>","DOI":"10.3390\/s24102993","type":"journal-article","created":{"date-parts":[[2024,5,8]],"date-time":"2024-05-08T12:27:11Z","timestamp":1715171231000},"page":"2993","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Deep Learning Based Over-the-Air Training of Wireless Communication Systems without Feedback"],"prefix":"10.3390","volume":"24","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4742-9635","authenticated-orcid":false,"given":"Christopher P.","family":"Davey","sequence":"first","affiliation":[{"name":"School of Mathematics, Physics and Computing, University of Southern Queensland, Springfield, QLD 4300, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5512-8843","authenticated-orcid":false,"given":"Ismail","family":"Shakeel","sequence":"additional","affiliation":[{"name":"Spectrum Warfare Branch, Information Sciences Division, Defence Science and Technology Group (DSTG), Edinburgh, SA 5111, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2290-6749","authenticated-orcid":false,"given":"Ravinesh C.","family":"Deo","sequence":"additional","affiliation":[{"name":"School of Mathematics, Physics and Computing, University of Southern Queensland, Springfield, QLD 4300, Australia"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4048-1676","authenticated-orcid":false,"given":"Sancho","family":"Salcedo-Sanz","sequence":"additional","affiliation":[{"name":"Department of Signal Processing and Communications, Universidad de Alcal\u00e1, 28805 Alcal\u00e1 de Henares, Madrid, Spain"}]}],"member":"1968","published-online":{"date-parts":[[2024,5,8]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"563","DOI":"10.1109\/TCCN.2017.2758370","article-title":"An Introduction to Deep Learning for the Physical Layer","volume":"3","author":"Hoydis","year":"2017","journal-title":"IEEE Trans. Cogn. Commun. Netw."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"2278","DOI":"10.1109\/LCOMM.2018.2868103","article-title":"Backpropagating through the Air: Deep Learning at Physical Layer without Channel Models","volume":"22","author":"Raj","year":"2018","journal-title":"IEEE Commun. Lett."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Aoudia, F.A., and Hoydis, J. (2018, January 28\u201331). End-to-End Learning of Communications Systems without a Channel Model. Proceedings of the 2018 52nd Asilomar Conference on Signals, Systems, and Computers, Grove, CA, USA.","DOI":"10.1109\/ACSSC.2018.8645416"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Ye, H., Li, G.Y., Juang, B.F., and Sivanesan, K. (2018, January 9\u201313). Channel Agnostic End-to-End Learning Based Communication Systems with Conditional GAN. Proceedings of the 2018 IEEE Globecom Workshops (GC Wkshps), Abu Dhabi, United Arab Emirates.","DOI":"10.1109\/GLOCOMW.2018.8644250"},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"O\u2019Shea, T.J., Roy, T., and West, N. (2019, January 18\u201321). Approximating the Void: Learning Stochastic Channel Models from Observation with Variational Generative Adversarial Networks. Proceedings of the 2019 International Conference on Computing, Networking and Communications (ICNC), Honolulu, HI, USA.","DOI":"10.1109\/ICCNC.2019.8685573"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"306","DOI":"10.1109\/TMC.2019.2950398","article-title":"Adversarial Deep Learning for Over-the-Air Spectrum Poisoning Attacks","volume":"20","author":"Sagduyu","year":"2021","journal-title":"IEEE Trans. Mob. Comput."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"D\u00f6rner, S., Henninger, M., Cammerer, S., and Brink, S.T. (2020, January 26\u201329). WGAN-based Autoencoder Training Over-the-air. Proceedings of the 2020 IEEE 21st International Workshop on Signal Processing Advances in Wireless Communications (SPAWC), Atlanta, GA, USA.","DOI":"10.1109\/SPAWC48557.2020.9154335"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Davey, C.P., Shakeel, I., Deo, R.C., and Salcedo-Sanz, S. (2023). Channel-Agnostic Training of Transmitter and Receiver for Wireless Communications. Sensors, 23.","DOI":"10.3390\/s23249848"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"132","DOI":"10.1109\/JSTSP.2017.2784180","article-title":"Deep Learning Based Communication Over the Air","volume":"12","author":"Cammerer","year":"2018","journal-title":"IEEE J. Sel. Top. Signal Process."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"5489","DOI":"10.1109\/TCOMM.2020.3002915","article-title":"Trainable Communication Systems: Concepts and Prototype","volume":"68","author":"Cammerer","year":"2020","journal-title":"IEEE Trans. Commun."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"3133","DOI":"10.1109\/TWC.2020.2970707","article-title":"Deep Learning-Based End-to-End Wireless Communication Systems with Conditional GANs as Unknown Channels","volume":"19","author":"Ye","year":"2020","journal-title":"IEEE Trans. Wirel. Commun."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"86834","DOI":"10.1109\/ACCESS.2022.3224922","article-title":"Innovative Variational AutoEncoder for an End-to-end Communication System","volume":"11","author":"Alawad","year":"2022","journal-title":"IEEE Access"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"631","DOI":"10.1109\/TCCN.2022.3161936","article-title":"Residual-Aided End-to-End Learning of Communication System without Known Channel","volume":"8","author":"Jiang","year":"2022","journal-title":"IEEE Trans. Cogn. Commun. Netw."},{"key":"ref_14","unstructured":"Kim, M., Fritschek, R., and Schaefer, R.F. (2023, January 27). Learning End-to-End Channel Coding with Diffusion Models. Proceedings of the WSA & SCC 2023; 26th International ITG Workshop on Smart Antennas and 13th Conference on Systems, Communications, and Coding, Braunschweig, Germany."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"2503","DOI":"10.1109\/JSAC.2019.2933891","article-title":"Model-free training of end-to-end communication systems","volume":"37","author":"Aoudia","year":"2019","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"ref_16","unstructured":"Goodfellow, I. (2016). Nips 2016 tutorial: Generative adversarial networks. arXiv."},{"key":"ref_17","unstructured":"Arjovsky, M., Chintala, S., and Bottou, L. (2017, January 6\u201311). Wasserstein generative adversarial networks. Proceedings of the International Conference on Machine Learning, PMLR, Sydney, NSW, Australia."},{"key":"ref_18","unstructured":"Bishop, C.M. (2023, October 10). Mixture Density Networks. Technical Report, Aston University. Available online: https:\/\/research.aston.ac.uk\/en\/publications\/mixture-density-networks."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"2703","DOI":"10.1021\/acsphotonics.0c00630","article-title":"Deep Convolutional Mixture Density Network for Inverse Design of Layered Photonic Structures","volume":"7","author":"Unni","year":"2020","journal-title":"ACS Photonics"},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Karoliny, J., Etzlinger, B., and Springer, A. (2020, January 7\u201311). Mixture Density Networks for WSN Localization. Proceedings of the 2020 IEEE International Conference on Communications Workshops (ICC Workshops), Dublin, Ireland.","DOI":"10.1109\/ICCWorkshops49005.2020.9145035"},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Mostafavi, S., Sharma, G.P., and Gross, J. (2023, January 4\u20138). Data-Driven Latency Probability Prediction for Wireless Networks: Focusing on Tail Probabilities. Proceedings of the GLOBECOM 2023\u20132023 IEEE Global Communications Conference, Kuala Lumpur, Malaysia.","DOI":"10.1109\/GLOBECOM54140.2023.10437281"},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Khurjekar, I.D., and Gerstoft, P. (2024, January 14\u201319). Multi-Source DOA Estimation With Statistical Coverage Guarantees. Proceedings of the ICASSP 2024\u20132024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Seoul, Republic of Korea.","DOI":"10.1109\/ICASSP48485.2024.10446097"},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"11117","DOI":"10.1109\/JSEN.2022.3223765","article-title":"Automotive Radar Modeling for Virtual Simulation Based on Mixture Density Network","volume":"23","author":"Li","year":"2023","journal-title":"IEEE Sensors J."},{"key":"ref_24","unstructured":"Ioffe, S., and Szegedy, C. (2015, January 6\u201311). Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift. Proceedings of the 32nd International Conference on Machine Learning, Lille, France."},{"key":"ref_25","unstructured":"Ramachandran, P., Zoph, B., and Le, Q.V. (2017). Searching for activation functions. arXiv."},{"key":"ref_26","first-page":"550","article-title":"Residual networks behave like ensembles of relatively shallow networks","volume":"29","author":"Veit","year":"2016","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1016\/j.neucom.2022.06.111","article-title":"A Comprehensive Survey and Performance Analysis of Activation Functions in Deep Learning","volume":"503","author":"Dubey","year":"2022","journal-title":"Neurocomputing"},{"key":"ref_28","unstructured":"Glorot, X., Bordes, A., and Bengio, Y. (2011, January 11\u201313). Deep sparse rectifier neural networks. Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, JMLR Workshop and Conference Proceedings, Fort Lauderdale, FL, USA."},{"key":"ref_29","first-page":"179","article-title":"Effects of HPA-nonlinearity on a 4-DPSK\/OFDM-signal for a digital sound broadcasting signal","volume":"332","author":"Rapp","year":"1991","journal-title":"ESA Spec. Publ."},{"key":"ref_30","unstructured":"Kingma, D.P., and Ba, J. (2014). Adam: A method for stochastic optimization. arXiv."},{"key":"ref_31","unstructured":"Izmailov, P., Podoprikhin, D., Garipov, T., Vetrov, D., and Wilson, A.G. (2018). Averaging weights leads to wider optima and better generalization. arXiv."},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Smith, L.N. (2017, January 24\u201331). Cyclical Learning Rates for Training Neural Networks. Proceedings of the 2017 IEEE Winter Conference on Applications of Computer Vision (WACV), Santa Rosa, CA, USA.","DOI":"10.1109\/WACV.2017.58"},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"22","DOI":"10.1109\/MRA.2020.2980591","article-title":"Movement primitive learning and generalization: Using mixture density networks","volume":"27","author":"Zhou","year":"2020","journal-title":"IEEE Robot. Autom. Mag."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"528","DOI":"10.1109\/TAI.2021.3108129","article-title":"Redefining Wireless Communication for 6G: Signal Processing Meets Deep Learning With Deep Unfolding","volume":"2","author":"Jagannath","year":"2021","journal-title":"IEEE Trans. Artif. Intell."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/24\/10\/2993\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T14:42:15Z","timestamp":1760107335000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/24\/10\/2993"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,5,8]]},"references-count":34,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2024,5]]}},"alternative-id":["s24102993"],"URL":"https:\/\/doi.org\/10.3390\/s24102993","relation":{},"ISSN":["1424-8220"],"issn-type":[{"type":"electronic","value":"1424-8220"}],"subject":[],"published":{"date-parts":[[2024,5,8]]}}}