Research Article
5G Cognitive Radio Networks Using Reliable Hybrid Deep Learning Based on Spectrum Sensing
Table 8
Performance analysis of the different algorithms at 52–2,200 MHz and low SNR (−20 db to 0 db).
| Sl. no. | Algorithm details | Samples ( samples) | Performance metrics | Sensing accuracy | Precision | Recall | | |
| 01 | LSTM-SS | | 0.965 | 0.955 | 0.950 | 0.1012 | 0.1002 | 02 | PALM-SS | 0.942 | 0.930 | 0.932 | 0.093 | 0.0923 | 03 | Hybrid ANN-SS | 0.921 | 0.910 | 0.923 | 0.089 | 0.087 | 04 | Normal energy detectors-SS | 0.782 | 0.762 | 0.780 | 0.086 | 0.0864 | 05 | RF-SS | 0.80 | 0.82 | 0.812 | 0.087 | 0.085 | 06 | Proposed scheme | 0.982 | 0.973 | 0.970 | 0.1012 | 0.1010 |
|
|