Multi-Level Fuzzy Inference System Based Handover Decision Model for Unmanned Vehicles

Abstract

In coming years, mobile communications systems will have a significant role for Unmanned Aerial Vehicles (UAVs), that will either function as mobile users or as mobile base stations in the sky. No doubt UAVs or drones provide a number of benefits for mobile communications networks and other non-communication services, but they also have a number of drawbacks, particularly when it comes to managing handover (HO). A number of approaches have already been suggested by various authors for making the HO easy and efficient, but only few parameters were considered for making the HO decision. However, with the advancement in technology and continuous rise in users, the handover system's reliance factors continue to grow, which could make the system more complex. As a result, the requirement for creating and implementing a system that can effectively manage system complexity arises. In order to reduce the complexity of the system, a novel method is proposed in this paper in which multi-level fuzzy system is used. The main goal of using the fuzzy system at different levels is to minimize the rule complexity of fuzzy systems at different levels which in turn enhances the performance of handover systems. In addition to this, number of parameters like coverage, speed limit, cost, connection time, security and power consumption were taken into consideration while designing the handover system. The model works in three levels, at the first level coverage, speed limit and cost parameters are processed to generate the first probability output. At the second level, connection time, security and power consumption parameters are considered to get the second probability output. The two outputs obtained thus serves as the input to third level where again they are processed to get the final output as estimation level. The efficacy of the suggested multi-level fuzzy system is analyzed in the MATLAB tool. The experimental results are obtained and compared with the traditional handover systems in terms of various dependency factors to prove its efficiency.