SVD Detection for Cognitive Radio Network based on Average of Maximum-Minimum of the ICDF

Spectrum sensing in cognitive radio (CR) has been a very important function to enable the state of the art technology in revolutionizing spectrum efficient utilization. SVD stands for Singular Value Decomposition. It's a method for matrix decomposition/factorization. The simplest way to visualize and understand how SVD is useful is to think in terms of Principal Component Analysis (PCA)/dimensionality reduction. In this paper we provide an average of maximum-minimum inverse cumulative distribution function (ICDF). We use raised cosine to test the performance of the signal detector to perform the simulation. The average eigenvalue based SVD signal detector was found to be more efficient in sensing signal without knowing the properties of the transmitted signal.