Table 3 The classic HMM and MOHT-HMM based spectrum prediction performance comparison.
\(L_1^{\text {ukn}}\) | \(L_2^{\text {ukn}}\) | \(L_3^{\text {ukn}}\) | \(L_4^{\text {ukn}}\) | \(L_5^{\text {ukn}}\) | \(L_6^{\text {ukn}}\) | \(L_7^{\text {ukn}}\) | |
|---|---|---|---|---|---|---|---|
HMM (measured data) | 0.8480 | 0.8586 | 0.8772 | 0.8996 | 0.9054 | 0.9280 | 0.9180 |
MOHT-HMM (estimated data) | 0.8365 | 0.8460 | 0.8550 | 0.8750 | 0.8860 | 0.8980 | 0.9086 |
\(L_8^{\text {ukn}}\) | \(L_9^{\text {ukn}}\) | \(L_{10}^{\text {ukn}}\) | \(L_{11}^{\text {ukn}}\) | \(L_{12}^{\text {ukn}}\) | \(L_{13}^{\text {ukn}}\) | \(L_{14}^{\text {ukn}}\) | |
|---|---|---|---|---|---|---|---|
HMM (measured data) | 0.9460 | 0.9368 | 0.9445 | 0.9280 | 0.8405 | 0.9022 | 0.8814 |
MOHT-HMM (estimated data) | 0.9390 | 0.9300 | 0.9230 | 0.9020 | 0.8249 | 0.8694 | 0.8498 |
