An Optimized Spectrum Sensing Implementation Based on SVM, KNN and TREE Algorithms

Cognitive radio (CR) network is an intelligent technology, widely used to solve the scarcity of the radio spectrum by allowing the unlicensed users to have access to the licensed spectrum. Spectrum sensing (SS) phase is of great importance to the workings of a cognitive radio network (CRN). It consists in detecting licensed signals in a particular frequency band to decide whether the unlicensed signals can transmit or not. In order to detect primary user (PU) presence, this paper proposes a low cost and low power consumption spectrum sensing implementation based on real signals. These signals are generated by an ARDUINO UNO card and a 433 MHz Wireless transmitter (ASK (Amplitude-Shift Keying) and FSK (Frequency-Shift Keying) modulation type). The reception interface is constructed using an RTL-SDR dongle connected to MATLAB software. The signal detection (spectrum sensing) is done by three methods: support vector machine (SVM), Decision Trees (TREE) and k-nearest neighbors (KNN). The main objective is to identify the best method for spectrum sensing between the three methods. The performance evaluation of our proposed model is the probability of detection (P_d) and the false alarm probability (P_fa). This Comparative work has shown that the SS operation by SVM and KNN can be more accurate than TREE and some other classical detectors.