QoE Handover Analysis in a Heterogeneous Network

Abstract

For the best mobility experience, users require a method that can receive services for a seamless handover. Handover occurs when a user device is in motion and requires continuity without any noticeable interruption in the service. Vertical handoff (VHO) technique is employed for this purpose to provide good connectivity to users within a wireless network. However, Quality of Service (QoS) cannot guarantee user satisfaction, a new approach is proposed called Quality of Experience to tackle this problem. QoE ensures that each user receives the service they wish, and their performance meets their expectations. Instead of focusing the quality the QoS follows, QoE focuses primarily on end-user satisfaction. Data of the satisfactory level of users is difficult to collect because it is time-consuming and is a very tedious process. Therefore, we use learning algorithms that can train and predict with the highest accuracy the satisfaction scores of the users. Random Neural Network can estimate the mean opinion score of users, without specifically collecting it from users.   Random Forest algorithm can predict and forecast the path of the user when establishing a handover and can easily store large data during a single handover. We compare both the algorithm and find the precision and conclude that Random Forest can predict with highest precision providing the best handover process.

Keywords: Random neural Network, Quality of Service, Quality of Experience, machine learning, handoff, heterogeneous network.