Dynamic Spectrum Sensing with Automatic Modulation Classification for a Cognitive Radio Enabled NomadicBTS
Folarin Olaloye, Emmanuel Adetiba · Journal of Physics Conference Series · 2019
An existing Nomadic Base Transceiver Station (NomadicBTS) architecture designed and implemented in the literature was built on software defined radio technology. This technology performs radio functions via software modules. Cognitive radio on the other hand is essentially built on software defined radio technology integrated with artificial intelligence. This research work extends the existing NomadicBTS architecture with cognitive radio capability for the purpose of introducing dynamic (opportunistic) spectrum sensing for efficient spectrum utilization in mobile networks. This is achieved by developing an Automatic Modulation Classification (AMC) model for spectrum sensing based on MultiLayer Perceptron (MLP) Artificial Neural Network (ANN). A suitable AMC model with optimum accuracy of 92.57% and Mean Square Error (MSE) of 0.0185 was empirically determined and deemed acceptable for incorporation into the NomadicBTS architecture.