Classification of IQ-Modulated Signals Based on Reservoir Computing With Narrowband Optoelectronic Oscillators

• Published in: IEEE journal of quantum electronics Vol. 57; no. 3; pp. 1 - 8
• Main Authors: Dai, Haoying, Chembo, Yanne K.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Classification; Computation; Continuous radiation; Hardware; IQ~modulation formats; Modulation; Narrowband; Neural networks; Noise generation; nonlinear oscillators; optoelectronic oscillators; Optoelectronics; Oscillators; Radio frequency; radio modulation recognition; Radio signals; Reservoir computing; Reservoirs; RF signals; Semiconductor lasers; Signal classification; Signal processing; Task analysis; Training; Wideband
• ISSN: 0018-9197; 1558-1713
• DOI: 10.1109/JQE.2021.3074132

We numerically perform the classification of IQ-modulated radiofrequency signals using reservoir computing based on narrowband optoelectronic oscillators (OEOs) driven by a continuous-wave semiconductor laser. In general, the OEOs used for reservoir computing are wideband and are processing analog signals in the baseband. However, their hardware architecture is inherently inadequate to directly process radiotelecom or radar signals, which are modulated carriers. On the other hand, the high-<inline-formula> <tex-math notation="LaTeX">Q </tex-math></inline-formula> OEOs that have been developed for ultra-low phase noise microwave generation have the adequate hardware architecture to process such multi-GHz modulated signals, but they have never been investigated as possible reservoir computing platforms. In this article, we show that these high-<inline-formula> <tex-math notation="LaTeX">Q </tex-math></inline-formula> OEOs are indeed suitable for reservoir computing with modulated carriers. Our dataset (DeepSig RadioML) is composed with 11 analog and digital formats of IQ-modulated radio signals (BPSK, QAM64, WBFM, etc.), and the task of the high-<inline-formula> <tex-math notation="LaTeX">Q </tex-math></inline-formula> OEO reservoir computer is to recognize and classify them. Our numerical simulations show that with a simpler architecture, a smaller training set, fewer nodes and fewer layers than their neural network counterparts, high-<inline-formula> <tex-math notation="LaTeX">Q </tex-math></inline-formula> OEO-based reservoir computers perform this classification task with an accuracy better than the state-of-the-art, for a wide range of parameters. We also investigate in detail the effects of reducing the size of the training sets on the classification performance.