An Autonomous Adaptive Enhancement Method Based on Learning to Optimize Heterogeneous Network Selection

Abstract

Mobile workstations are frequently used in heterogeneous network's challenging environments. Users must move between various networks for a myriad of purposes, including vertical handover. At this time, it is critical for the mobile station to quickly pick the most appropriate networks from all identified alternative connections with the decision outcome, avoiding the ping-pong effect to the greatest extent feasible. Based on a combination of network characteristics as well as user choice, this study offers a heterogeneous network selection method. This technique integrates three common Multi-Attribute Decision-Making (MADM) techniques, notably the Fuzzy Analytic Hierarchy Process (FAHP), Entropy, and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), to take into consideration user preferences for every prospective network as well as the real scenario of heterogeneous networks. For different traffic classes, FAHP is first utilized to determine the weights of network parameters and the utility numbers of total options available. Next, entropies and TOPSIS are utilized to obtain only the unbiased weightages of network factors and utility principles of totally different options. The most suitable networks, whose utility number is the greatest and larger than that of the equivalent number of present networks of the phone station, are chosen to provide accessibility based on the utility numbers of each prospective system as a limit. The suggested method not only eliminates a particular algorithm's one-sided character but also dynamically changes the percentage of each method in the desired outcome based on real needs. The proposed model was compared to the three existing hybrid methods. The results showed that it could precisely choose the optimized network connectivity and significantly reduce the value of vertical handoffs. It also provides the requisite Quality of Service (QoS) and Quality of Everything (QoE) in terms of the quantitative benefits of vertical handovers.