Perbandingan Estimasi Derau tanpa Informasi Sinyal Transmisi dengan Masukan Sinyal DVB-T pada Sistem Radio Kognitif

Abstrak

The comparison of the performance of three algorithms of blind noise estimation without a priori information about transmitted signal is conducted in this research. The three estimators are maximum likelihood estimator, estimator based on cyclic prefix and estimator based on Forward consecutive mean excision (FCME). The comparison is done with the input DVB-T signal and with varied SNR and time observation. The performance of the three estimators are measured with NMSE. Cyclic prefix-estimator and Maximum likelihood-estimator has a stable performance against SNR, while FCME estimator has a fluctuative result. The number of symbol has an impact for three estimators, thus the performances are getting better as the number of symbol is increasing. The comparison shows that maximum likelihood-estimator has the best performances of all. It means that this estimator has high accuracy.