| (K, N) | M | 链路平均频谱效率(bps/Hz) | 强化学习方法 输出层大小 | 平均 迭代次数 | ||||||
| 强化学习方法 | 其他学习方案 | |||||||||
| Proposed | ER | Joint | Ideal FP | Delayed FP | Random | Proposed | Joint | FP | ||
| (5, 20) | 1 | 1.59 | 1.58 | 1.60 | 1.56 | 1.46 | 0.41 | 1 + 1 | 10 | 70.30 |
| 2 | 2.80 | 2.79 | 2.79 | 2.68 | 2.53 | 0.99 | 2 + 1 | 20 | 105.73 | |
| 4 | 4.59 | 4.56 | 4.49 | 3.71 | 3.54 | 2.11 | 4 + 1 | 40 | 121.56 | |
| (5, 20) | 6 | 5.92 | 5.86 | 5.52 | 4.47 | 4.25 | 3.04 | 6 + 1 | 60 | 135.08 |
| 8 | 6.89 | 6.64 | 5.64 | 5.06 | 4.83 | 3.81 | 8 + 1 | 80 | 142.30 | |
| 10 | 7.05 | 6.97 | 5.71 | 5.44 | 5.02 | 4.46 | 10 + 1 | 100 | 156.86 | |