Semi-supervised Learning for Automatic Modulation Recognition Using Haar Time–Frequency Mask and Positional–Spatial Attention

<div>Confusion matrix of HTF-PSA-SSL under (a) −6, (b) 0, and (c) 12 dB signal. The horizontal axis is the predicted modulation type of network. The vertical axis is the true modulation type of signals. Each number on the leading diagonal shows the total number of modulation correct prediction.</div>

Wireless Communications and Mobile Computing

fig7

Figure 7

Figure 7: Semi-supervised Learning for Automatic Modulation Recognition Using Haar Time–Frequency Mask and Positional–Spatial Attention 