Analysis of the mathematical method of detection SVD in cognitive radio

Abstract

In this work, the performance of the Singular Value Decomposition method (SVD) applied to the users detection in a cognitive mobile network is evaluated. The performance of the SVD method was analyzed in terms of probability of detection (Pd) vs signal-to-noise ratio (SNR), obtaining a mathematical model based on an accumulated probability distribution (CDF). The results were compared with the CDFs of the theoretical SVD and energy detection methods, using the statistical maximum likelihood estimator (MLE), concluding that the method applied in the evaluated system exceeded the theoretical methods in terms of Pd.

Keywords: Cognitive Radio in Mobile Networks, Decomposition into Singular Values (SVD), Detection Probability (Pd), Maximum Verosimilitude Estimator (MLE), Accumulated Probability Distribution (CDF).