We propose a QAM demodulator based on the AWSOM (All Winner SOM), a clustering algorithm optimized for efficient hardware implementation. The demodulator has been modeled and simulated in a MATLAB/SIMULINK environment, and the Machine Learning algorithm has been implemented on FPGA. The simulation results show its ability to demodulate data also in the presence of channel noise, phase error, and nonlinear distortions. The data is given in terms of demodulation capabilities and confusion matrix for the recognition of symbols. In the worst case, where we considered an SNR of 5dB, the symbol error-rate of our AW-SOM approach is only 5.28%. We show how our system is a viable alternative to traditional SOM for the above-mentioned application due to its clustering performance and its efficiency if implemented in hardware, as proved by the comparison with the state of the art.

Canese, L., Cardarilli, G.c., Di Nunzio, L., Fazzolari, R., Re, M., Spano, S. (2023). A Hardware-Oriented QAM Demodulation Method Driven by AW-SOM Machine Learning. In Conference record of the fifty-seventh Asilomar Conference on Signals, Systems & Computers (pp.937-941). IEEE Computer Society [10.1109/IEEECONF59524.2023.10476985].

A Hardware-Oriented QAM Demodulation Method Driven by AW-SOM Machine Learning

Canese L.;Cardarilli G. C.;Di Nunzio L.;Fazzolari R.;Re M.;Spano S.
2023-01-01

Abstract

We propose a QAM demodulator based on the AWSOM (All Winner SOM), a clustering algorithm optimized for efficient hardware implementation. The demodulator has been modeled and simulated in a MATLAB/SIMULINK environment, and the Machine Learning algorithm has been implemented on FPGA. The simulation results show its ability to demodulate data also in the presence of channel noise, phase error, and nonlinear distortions. The data is given in terms of demodulation capabilities and confusion matrix for the recognition of symbols. In the worst case, where we considered an SNR of 5dB, the symbol error-rate of our AW-SOM approach is only 5.28%. We show how our system is a viable alternative to traditional SOM for the above-mentioned application due to its clustering performance and its efficiency if implemented in hardware, as proved by the comparison with the state of the art.
57th Asilomar Conference on Signals, Systems and Computers, ACSSC 2023
Asilomar Hotel and Conference Grounds, Pacific Grove, CA, USA
2023
57
Rilevanza internazionale
2023
Settore ING-INF/01
English
Demodulation
Machine learning
Self organizing maps
Telecommunications
Intervento a convegno
Canese, L., Cardarilli, G.c., Di Nunzio, L., Fazzolari, R., Re, M., Spano, S. (2023). A Hardware-Oriented QAM Demodulation Method Driven by AW-SOM Machine Learning. In Conference record of the fifty-seventh Asilomar Conference on Signals, Systems & Computers (pp.937-941). IEEE Computer Society [10.1109/IEEECONF59524.2023.10476985].
Canese, L; Cardarilli, Gc; Di Nunzio, L; Fazzolari, R; Re, M; Spano, S
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2108/364064
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact