Simulation 2

This simulation assumed that f₁(t) was an AM complex tone the same as Fig. 8 and that f₂(t) was another AM complex tone, where F₀(t)=300 Hz, N_F0=10, and the tone's instantaneous amplitude was sinusoidal (15 Hz). Therefore, harmonics of f₁(t) and f₂(t) in multiples of 600 Hz (for example, the third harmonic of f₁(t) and second harmonic of f₂(t)) exist in the same frequency region. Five types of f(t) were used as simulation stimuli, where the SNRs of f(t) ranged from 0 to 20 dB in 5-dB steps.

For example, when the SNR of f(t) was 10 dB, as shown in Fig. 12, the proposed method could segregate A_k(t) with high accuracy and could extract $\hat{f}_1(t)$, shown in Fig. 13, from the f(t), even when two components of the signals existed in the same frequency region. In this case, the precision for A_k(t) is shown in Fig. 14 (top panel). In addition, the average SDs of $\hat{f}_1(t)$ and f(t) for five simulations are shown in Fig. 14 (bottom panel). It was possible to improve the precision by about 3.1 dB and the spectrum distortion by about 7.2 dB as noise reduction, comparing the proposed method with condition 3.

Hence, just like the results of the previous simulations, the proposed model could also extract with high precision the amplitude information of signal f₁(t) from a noise-added signal f(t) in which two AM complex tones existed in the same frequency region. Here, comparing the results of the proposed method with condition 2, we see that the segregated accuracy using the proposed method shoed more improvement with simulation 2.