QoE-based vertical handover decision management for cognitive networks using ANN

Abstract

The main goal of 5G cognitive radio in heterogeneous networks is to maintain seamless connectivity and provide satisfying Quality of Service (QoS) by switching from one network to another using Vertical Handovers (VHO). The accuracy and the quickness of VHO decision is the key feature to improve and maintain high QoS levels. Recently, many works have been interested to Artificial Neural Networks (ANN) as decisional algorithm for VHO. It was considered efficient since it can handle a decision based on prior knowledge acquired during a learning process and takes into account many criteria of QoS. This efficiency can be more improved while considering the Quality of Experience (QoE) in the ANN inputs set. This paper proposes a modified multi-criteria vertical handoff decision algorithm based on ANN with QoE prediction scheme. The proposed mechanism concept serves to improve the accuracy of the vertical handoff decision for radio heterogeneous networks. Developed algorithm is compared to classical Fuzzy Logic (FL) and Multi Attribute Decision-Making (MADM) ones. Obtained results show that QoE based ANN improve final QoS/QoE satisfaction metrics while reducing delays and the number of executed handoffs.