Читать книгу Handbook of Intelligent Computing and Optimization for Sustainable Development - Группа авторов - Страница 133

In this chapter, we have presented the ML/DL application for wireless communications systems. A brief overview of the ML algorithms applied to some of the emerging wireless technologies is explained. The key technologies of the 5G network were discussed in detail. The utility of the current DL architectures to train and predict the wireless signal classification is presented. We have also proposed a CNN model for AMC and compared with the existing model. The implementation results show that the proposed model outperforms the existing one. Further, modified InceptNet was proposed for CSI feedback for FDD massive MIMO system. The proposed InceptNet compared with the prevailing CsiNet attains superior results in terms of NMSE and cosine similarity. Both the models were tested for different compression ratios, and it is observed that for all the values of CR, InceptNet surpasses the CsiNet. The chapter briefly discussed the future research directions and critical enabling technologies for next-generation (5G and Beyond) wireless networks for ML applications.