Multiband Cooperative Spectrum Sensing Meets Vehicular Network: Relying on CNN-LSTM Approach

A vehicular network is expected to empower all aspects of the intelligent transportation system (ITS) and aim at improving road safety traffic efficiency. In view fact that spectrum scarcity becomes more severe owing increasing number connected vehicles, implying sensing technology in network, i.e., cognitive has emerged as a promising solution provide opportunistic usage licensed spectrum. However, some unique features networks, such high movement dynamic topology, take on challenges for sensing. Recently, machine learning-based approaches, especially deep learning, have attracted sufficient interest. this work, we investigate cooperative (CSS) approach multiband network. Specifically, integrate two powerful learning models, convolutional neural (CNN) exploit from data, long-short-term memory (LSTM) then utilized extract temporal correlations given input generated by CNN structure. Instead predefined decision threshold typically set conventional our proposed could eliminate impact impertinent value setting. Extensive simulations been conducted evaluate effectiveness method achieving satisfactory performance, particularly terms higher detection accuracy, robustness low signal-to-noise ratio (SNR) environments, significant reduction time compared other methods.