Abstract
The forthcoming fifth-generation networks require improvements in cognitive radio intelligence, going towards more smart and aware radio systems. In the emerging radio intelligence approach, the empowerment of cognitive capabilities is performed through the adoption of machine learning techniques. This paper investigates the combined application of the convolutional and recurrent neural networks for the channel state information forecasting, providing a multivariate scalar time series prediction by taking into account the multiple factors dependence of the channel state conditions. Finally, the system performance has been analyzed in terms of prediction accuracy expressed as absolute deviation error and mean percentage error, in comparison with an alternative machine learning method recently proposed in the literature with the aim at solving the same prediction problem.
Highlights
The emergence of the need for new network technologies and architectures has been launched by the diffusion of novel services and applications demanding high data rates and reliability, low latency, and congestion levels
The aims of the paper are summarized as Design and development of an machine learning (ML)-based framework involving the usage of the convolutional neural networks (CNNs) and the recurrent neural network (RNN) to predict the channel state information (CSI) behavior; Extensive numerical simulations to provide the performance evaluation analysis as regards the prediction accuracy expressed as mean absolute deviation error and mean percentage error between the actual CSI value and those predicted; Testing of the proposed framework under different application scenarios, i.e., outdoor and indoor conditions, in comparison with the deep learning strategy proposed in [2] and that presented in [16]
This paper has dealt with the CSI prediction problem, by proposing a forecasting strategy combining both the CNNs and the RNNs approaches and providing a method for predicting the CSI
Summary
The emergence of the need for new network technologies and architectures has been launched by the diffusion of novel services and applications demanding high data rates and reliability, low latency, and congestion levels. Cognitive radio (CR) approaches have played a crucial role, aiming at designing context-driven strategies. In this regard, the advent of the 5G networks, which involve the usage of the millimeter wave (mmWave) band has resulted in a large variety of obstacles such as pathloss, blockage, and high oxygen absorption [1]. The advent of the 5G networks, which involve the usage of the millimeter wave (mmWave) band has resulted in a large variety of obstacles such as pathloss, blockage, and high oxygen absorption [1] All these factors increase the complexity of detecting actual channel conditions and signals characteristics. An accurate value of the CSI plays a role of paramount importance in a wide range of practical applications, while inaccuracy on the CSI prediction value leads inevitably to substantial degradation on communication effectiveness
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
