Abstract:Cognitive radio (CR) has become an interesting research field that attracts researchers due to its increasing spectrum efficiency. Therefore, spectrum sensing (SS) is the essential function of cognitive radio systems. This paper presents an efficient SS model based on convolutional neural networks (CNNs). We use the spectrogram images of the received signals as the input to the CNN and use various images for signal and noise at different low primary‐user (PU) signal‐to‐noise ratios (SNRs) to train the network model. The model extracts the main features from the spectrogram images to represent signals and noise. Hence, this model can efficiently discriminate between signal and noise at different SNRs. The detection performance of the suggested model is compared with those of the traditional one‐stage, two‐stage SS methods, and different previous CNN models. The obtained outcomes demonstrate that the suggested model increases the detection accuracy more than those of the previous one‐stage SS methods by 17% at low SNRs of −20 dB and more than the previous two‐stage SS methods by 8% at low SNRs of −20 dB. In addition, it is demonstrated that the suggested model offers shorter sensing times than those of the two‐stage and one‐stage SS methods in the orders of 16.3, 16.6, 1.1, and 1.5 ms at SNRs of −20, −15, −10 and −5 dB, respectively. Furthermore, the proposed model improves the detection accuracy better than the different previously compared CNN models by 28% and 19% at low SNRs of −20 and −15 dB, respectively. This paper presents an efficient SS model based on CNNs. The detection performance of the suggested model is compared with those of the traditional one‐stage, two‐stage, and previous CNN models. The obtained outcomes demonstrate that the suggested model increases the detection accuracy more than those of the previous one‐stage SS methods by 17% at low SNRs of −20 dB and more than the previous two‐stage SS methods by 8% at low SNRs of −20 dB. In addition, it is demonstrated that the suggested model offers shorter sensing times than those of the two‐stage and one‐stage SS methods in the orders of 16.3, 16.6, 1.1, and 1.5 ms at SNRs of −20, −15, −10, and −5 dB, respectively. Furthermore, the proposed model improves the detection accuracy better than the different previously compared CNN models by 28% and 19% at low SNRs of −20 and −15 dB, respectively.

Comparison of Neural Network Architectures for Spectrum Sensing

Multi-Task Convolutional Neural Network Based Signal Recognition in Radio-over-Fiber Systems

Spectrum Analysis and Convolutional Neural Network for Automatic Modulation Recognition

Residual Network Structure-based High Accuracy Spectral Analysis Method

Deep CM-CNN for Spectrum Sensing in Cognitive Radio

Communication spectrum prediction method based on convolutional gated recurrent unit network

Spectrum sensing for cognitive radio based on convolution neural network

Deep Neural Networks for Spectrum Sensing: A Review

Spectrum Sensing for Symmetric α-Stable Noise Model With Convolutional Neural Networks

Convolutional neural network model for spectrum sensing in cognitive radio systems

Sequential Convolutional Recurrent Neural Networks for Fast Automatic Modulation Classification

Spatio-temporal Spectrum Load Prediction using Convolutional Neural Network and ResNet

A Neural Network Detector for Spectrum Sensing under Uncertainties

Spectrum Sensing Method Based on Residual Cellular Network

RamanNet: A generalized neural network architecture for Raman Spectrum Analysis

Spectrum Sensing Method Based on STFT-RADN in Cognitive Radio Networks

RL-Based Hyperparameter Selection for Spectrum Sensing With CNNs

Deep learning-based spectrum sensing and modulation categorization for efficient data transmission in Cognitive Radio

A Deep-Learning-Based Method for Spectrum Sensing with Multiple Feature Combination

1-D Convolutional Neural Network-Based Models for Cooperative Spectrum Sensing

Cooperative Spectrum Sensing Based on LSTM-CNN Combination Network in Cognitive Radio System