Leveraging Deep Learning and CSI for Handover in Wireless Network

Machine Learning and Deep Learning Models are being continuously be employed for network optimization and handover initialization in wireless networks, due to their capability to handle copious amounts of data. As the number of users and multimedia applications rises, bandwidth efficiency in cellular networks has emerged as a crucial factor in system design. Bandwidth is an essential resource utilised by wireless networks. Therefore, it is essential to improve bandwidth efficiency. Orthogonal Frequency Division Multiplexing (OFDM) and Non-Orthogonal Multiple Access (NOMA) are the primary candidates for contemporary wireless networks. NOMA is a method that segregates data from numerous users within the power domain. The suggested method introduces a machine learning-based handover mechanism between OFDM and NOMA, contingent upon channel conditions. The criterion for switching or handover has been established as the system's BER. A comparison analysis with previous research demonstrates that the suggested approach surpasses current techniques regarding SNR requirements and well as accuracy, hence enhancing the system's practical use under fading channel conditions.