Abstract
Blind modulation classification is an important step in implementing cognitive radio networks. The multiple-input multiple-output (MIMO) technique is widely used in military and civil communication systems. Due to the lack of prior information about channel parameters and the overlapping of signals in MIMO systems, the traditional likelihood-based and feature-based approaches cannot be applied in these scenarios directly. Hence, in this paper, to resolve the problem of blind modulation classification in MIMO systems, the time–frequency analysis method based on the windowed short-time Fourier transform was used to analyze the time–frequency characteristics of time-domain modulated signals. Then, the extracted time–frequency characteristics are converted into red–green–blue (RGB) spectrogram images, and the convolutional neural network based on transfer learning was applied to classify the modulation types according to the RGB spectrogram images. Finally, a decision fusion module was used to fuse the classification results of all the receiving antennas. Through simulations, we analyzed the classification performance at different signal-to-noise ratios (SNRs); the results indicate that, for the single-input single-output (SISO) network, our proposed scheme can achieve 92.37% and 99.12% average classification accuracy at SNRs of −4 and 10 dB, respectively. For the MIMO network, our scheme achieves 80.42% and 87.92% average classification accuracy at −4 and 10 dB, respectively. The proposed method greatly improves the accuracy of modulation classification in MIMO networks.
Highlights
The increase in communication demands and the shortage of spectrum resources have caused the cognitive radio (CR) and multiple-input multiple-output (MIMO) techniques to be implemented in wireless communication systems
Rajendran et al [15] proposed a new data-driven model for blind modulation classification (BMC) based on long short-term memory (LSTM), which learned the features from the time-domain amplitude and phase information of the modulation schemes and yielded an average classification accuracy close to 90% for signal-to-noise ratios (SNRs) from 0 to 20 dB
In order to verify the performance of the proposed scheme, some benchmark schemes are introduced, such as the SqueezeNet-based method [43], the GoogleNet-based method [44], the scheme based on the smooth pseudo Wigner–Ville distribution (SPWVD) proposed in [7], and the scheme based on the Wigner–Ville distribution (WVD) proposed in [31]
Summary
The increase in communication demands and the shortage of spectrum resources have caused the cognitive radio (CR) and multiple-input multiple-output (MIMO) techniques to be implemented in wireless communication systems. From this table, we can note that most of the proposed approaches are FB-based; concretely, the authors in [20,21] proposed similar methods for the MC of MIMO transceiver systems that calculate the higher-order statistical moments and cumulants of the received signals. More interesting studies about this topic concern the modulation classification for MIMO orthogonal frequency division multiplexing (OFDM) systems, as the MIMO OFDM has been widely adopted by many commercial standards, such as LTE and the Wifi For this problem, different approaches, such as the approximate Bayesian inference method, the Gibbs sampling-based method, and the joint independent component analysis (ICA). All these studies are traditional feature-based or likelihood-based approaches
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
