Fig. 1. DFT-S-OFDM system based on Gaussian mixture model

Fig. 2. 16QAM constellation of received single carrier signal with $\beta =\mathrm{3}$ and $E_{b(\mathrm{o}\mathrm{p}\mathrm{t})}/N_{\mathrm{0}}=\mathrm{30}\mathrm{d}\mathrm{B}$

Fig. 3. Performance of 16QAM DFT-S-OFDM system in nonlinear channel. (a) BER performance of 16QAM DFT-S-OFDM system with different $\beta$ values; (b) 16QAM GMM cluster demodulation constellation with $\beta =\mathrm{5}\mathrm{d}\mathrm{B}$,$E_{b(\mathrm{o}\mathrm{p}\mathrm{t})}/N_{\mathrm{0}}=\mathrm{16.8}\mathrm{d}\mathrm{B}$ at Ⅰ in Fig. 3(a); (c) 16QAM GMM cluster demodulation constellation with $\beta =\mathrm{4}\mathrm{d}\mathrm{B}$, $E_{b(\mathrm{o}\mathrm{p}\mathrm{t})}/N_{\mathrm{0}}=\mathrm{18}\mathrm{d}\mathrm{B}$ at Ⅱ in Fig. 3(a); (d) 16QAM GMM cluster demodulation constellation with $\beta =\mathrm{3}\mathrm{d}\mathrm{B}$, $E_{b(\mathrm{o}\mathrm{p}\mathrm{t})}/N_{\mathrm{0}}=\mathrm{21}\mathrm{d}\mathrm{B}$ at Ⅲ in Fig. 3(a)

Fig. 4. Performance of DFT-S-OFDM system in nonlinear channel. (a) BER performance of 32QAM DFT-S-OFDM system with different $\beta$ values; (b) BER performance of 64QAM DFT-S-OFDM system with different $\beta$ values

Fig. 5. SNR gain of DFT-S-OFDM system under different modulation orders

Table 1. Simulation parameters of DFT-S-OFDM system based on GMM cluster demodulation