Abstract
The task of equalization on the receiver side of a wireless communication system is typically accomplished with model-based estimation methods. However, the utilization of data-driven approaches, e.g., neural networks (NNs), for equalization is in focus of current research. In this work, we investigate two different NNs for single-carrier frequency domain equalization. We elaborate on how existing model knowledge can be incorporated into NNs, we introduce a data normalization scheme required for the regarded NNs, and we compare these data-driven methods with model-based approaches concerning performance and complexity.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We refer to an NN whose structure is deduced from model knowledge as “model-inspired”.
- 2.
Let \(\textbf{v}_{\text {in}}^{(q)}\) and \(\textbf{v}_{\text {hid}}^{(q)}\) denote the input and the output of the qth hidden layer. When employing a weighted residual connection, the input \(\textbf{v}_{\text {in}}^{(q+1)}\) of the follow-up hidden layer is \(\textbf{v}_{\text {in}}^{(q+1)} = \beta \textbf{v}_{\text {in}}^{(q)} + (1-\beta )\textbf{v}_{\text {hid}}^{(q)}\), with \(\beta \in [0,1[\).
- 3.
The complexity analysis is conducted in the same way as in [1].
References
Baumgartner, S., Bognár, G., Lang, O., Huemer, M.: Neural network based data estimation for unique word OFDM. In: Proceedings of the 55th Asilomar Conference on Signals, Systems, and Computers, pp. 381–388 (2021)
Fakatselis, J.: Criteria for 2.4 GHz PHY Comparison of Modulation. IEEE Document (1997), IEEE P802.11-97/157r1
Hershey, J.R., Le Roux, J., Weninger, F.: Deep unfolding: model-based inspiration of novel deep architectures. arXiv preprint arXiv:1409.2574 (2014)
Huemer, M., Koppler, A., Weigel, R., Reindl, L.: A review of cyclically extended single carrier transmission with frequency domain equalization for broadband wireless transmission. Eur. Trans. Telecommun. 14(4), 329–341 (2003)
Kay, S.M.: Fundamentals of Statistical Signal Processing: Estimation Theory, vol. 1. Prentice Hall, Hoboken (1993)
Klambauer, G., Unterthiner, T., Mayr, A., Hochreiter, S.: Self-normalizing neural networks. In: Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS), pp. 972–981 (2017)
O’Shea, T., Hoydis, J.: An introduction to deep learning for the physical layer. IEEE Trans. Cogn. Commun. Netw. 3(4), 563–575 (2017)
Reinhardt, S., Buzid, T., Huemer, M.: Receiver structures for MIMO-SC/FDE systems. In: Proceedings of the 2006 IEEE 63rd Vehicular Technology Conference, pp. 1401–1405 (2006)
Samuel, N., Diskin, T., Wiesel, A.: Learning to detect. IEEE Trans. Sig. Process. 67(10), 2554–2564 (2019)
Szegedy, C., et al.: Going deeper with convolutions. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1–9 (2015)
Witschnig, H., Mayer, T., Springer, A., Maurer, L., Huemer, M., Weigel, R.: The advantages of a Known Sequence versus cyclic Prefix in a SC/FDE system. In: Proceedings of the 5th International Symposium on Wireless Personal Multimedia Communications, vol. 3, pp. 1328–1332 (2002)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Baumgartner, S., Lang, O., Huemer, M. (2022). Neural Network Based Single-Carrier Frequency Domain Equalization. In: Moreno-Díaz, R., Pichler, F., Quesada-Arencibia, A. (eds) Computer Aided Systems Theory – EUROCAST 2022. EUROCAST 2022. Lecture Notes in Computer Science, vol 13789. Springer, Cham. https://doi.org/10.1007/978-3-031-25312-6_34
Download citation
DOI: https://doi.org/10.1007/978-3-031-25312-6_34
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-25311-9
Online ISBN: 978-3-031-25312-6
eBook Packages: Computer ScienceComputer Science (R0)