Narrow Results

Detection of relationship between two time series is so important in different scientific fields. Most common techniques are usually sensitive to stationarity or normality assumptions. In this research, a new copula-based method (cyclocopula) is introduced to detect the relationship between two cylostationary time series with fractional Brownian motion (fBm) errors. The performance of the proposed method is studied by employing numerous simulated datasets. The applicability of the introduced approach is also investigated in real-world problems. The numerical and applied studies verify the performance of the introduced technique.

To solve the low detection rate of the primary user in the cognitive radio environment, we propose a spectrum sensing method based on AdaBoost in the case of low SNR. In this paper, a set of received signal spectrum features are first calculated and extracted the discriminant feature vector as training samples and testing samples for classification. Finally, we utilize the trained AdaBoost to detect the primary user. Test result shows that the proposed algorithm is not affected by uncertainty factors of noise and has high performance to classification detection compared with ANN, SVM and maximum-minimum eigenvalue (MME).