A Data Driven Automated Handover Mechanism For Iot Networks

Abstract

With the increase in data traffic in IoT networks, handover in networks has become imminent as a single multiple access technique may not perform satisfactorily under all conditions. Therefore, selecting efficient multiple access mechanisms is essential to optimise the use of available bandwidth. It is important to implement a switching or handover mechanism to ensure excellent service quality, capable of selecting the most efficient multiple access approach according to channel conditions. Given the characteristics of wireless networks, it is essential to continuously assess them and choose the most suitable technique to achieve optimal Quality of Service (QoS). Advancements in machine learning and deep learning have enabled more efficient predictions of channel state information, resulting in improved system performance metrics. This research introduces a machine learning-augmented handover methodology for software-defined networks in next-generation wireless networks. This study proposes the simultaneous sensing of primary and secondary multiple access techniques for wireless networks, with a transition to the one offering superior QoS. Nevertheless, the principal access method must be prioritised to enhance system efficiency. The simulations conducted account for the path loss factor and SINR conditions in a realistic network setting.