Enhancing 5G-4G handover decisions with fuzzy inference systems

Abstract Handover mechanisms are important to ensure continuity of services in wireless networks. This study examines how Fuzzy Logic can be used in conjunction with the Adaptive Neuro-Fuzzy Inference System (ANFIS) to optimize the vertical handover decisions and network choices in mobile networks. The methodology is designed in a three-level architecture: initial handover determination, assessment of target availability and final network selection. Unlike prior single-stage FIS or ANFIS approaches, this three-stage cascaded pipeline jointly optimizes handover triggering, target availability estimation, and network selection using several heterogeneous QoS and mobility parameters within a unified framework. An ANFIS model based on a Sugeno fuzzy structure is utilized in order to guarantee flexibility and efficient learning. This system is trained through a hybrid algorithm that is aimed at controlling the epoch cycles and reducing errors. The proposed framework is clearly operationally superior to the one that was validated by MATLAB simulations. In particular, the ANFIS-based system improves the decision-making process that the traditional fuzzy logic models have a 16.3 to 35.86 margin of improvement at the range of 0.73 to 0.77 of availability metrics.