An End-to-End Reconfiguration System-Friendly Vertical Handover Scheme

Relaying information without accurately understanding the resource and service needs of the newly relocated node is a major obstacle to providing a smooth communication system during vertical handover in heterogeneous networks. There may be a lack of study on the topic of smooth vertical handover, according to the literature review. This research proposes a new technique for vertical handover in heterogeneous integrated networks that uses multi-attribute and neural networks. By navigating among UMTS, GPRS, WLAN, 4G, and 5G, we build the algorithm's framework according to the network environment. In order to set the input layer neurons of the BP neural network and train and learn from subsequent data, each network constructs a unique Q-Learning (Reinforcement Learning) model. Reference objects used in this process include the highest possible transmission rate, the minimum delay, SINR (signal to noise and interference ratio), bit error rate, user moving velocity, and packet loss rate, all of which can impact the wireless network's overall performance. In the end, the five wireless connections are evaluated based on the network download rate, which is used as a forecast goal. Then, the algorithm for vertical handover will choose the correct wireless network to make the choice. This research presents a vertical handover method that, according to MATLAB simulation results, enables efficient handover with seamless connection across multi-heterogeneous networks with a handover success rate of up to 90%.