Esquemas Caos-NOMA Baseados no Atrator de Lorenz

In this work, we propose a non-orthogonal multiple access (NOMA) scheme where the chaotic signals transmitted by each user are generated by the Lorenz chaotic attractor, joining the spectral efficiency of the NOMA scheme with the dynamical properties of chaotic signals, which we denominate chaos-NOMA. We employ a principal component analysis (PCA) algorithm for dimensionality reduction and obtain an orthonormal basis for the Lorenz attractor. Using this basis, we design a signal constellation for the chaotic signals generated by the attractor. The dynamical properties of the chaotic signals result in the transmission of time-variant waveforms. This variation is modeled as a intrinsic noise, which depending on the number of users and the power difference of the transmitted signals can result in a performance curve with error floor. We propose a neural network coupled to the demodulator to mitigate the effect of this noise.