A multicriteria decision-making framework for Hybrid Light Fidelity network

Demand for faster internet communications has been significantly on the rise, this is apparent given available spectrum scarcity and increasing demand for wireless services which leads to spectrum crunch problem. It occurs when more devices rely on wireless communication, overloading the limited...

Full description

Bibliographic Details
Main Author: Abdal, Rozin Badeel
Format: Thesis
Language:English
Published: 2024
Subjects:
Online Access:http://psasir.upm.edu.my/id/eprint/120084/
http://psasir.upm.edu.my/id/eprint/120084/1/120084.pdf
Description
Summary:Demand for faster internet communications has been significantly on the rise, this is apparent given available spectrum scarcity and increasing demand for wireless services which leads to spectrum crunch problem. It occurs when more devices rely on wireless communication, overloading the limited spectrum resources. Accordngly, this causes slower data rates and performance degradation. To fulfil the growing need for a better wireless connection, Light Fidelity (Li-Fi) has emerged as a solution. Unlike radiobased communication systems, Li-Fi transmits data by Visible Light rather than Radio Frequencies. Li-Fi has various benefits, including quicker data transfer, more bandwidth, less network congestion and interference. Li-Fi has numerous advantages, including but not limited to faster data transfer rates, more bandwidth and reduced network congestion, in addition to zero interference. Combining both technologies Wi-Fi and Li-Fi offers various benefits, including increased data transfer speed, and enhanced network and user capacity. Despite the hybrid approach suitability and strength, it still faces problems when encountering situations with Access Point Assignment (APA), Network Selection (NS), load balancing, and Handover (HO) problems. All these problems affect user connections in terms of maintaining a stable connection, specifically, HO leads to connection disruption of user from the access point (AP), the APA problem refers to choosing most suitable AP based on multiple factors, and balancing the load for connected users is challenging. The fluctuation of different variables values complicates the decision-making process in the midst of the three problems. Therefore, considering multi criteria measurement and evaluation will is anticipated to offer more robust decisionmaking mechanism which can assist users in addressing such problems. Different attempts were made to address these issues, where fuzzy logic, and load balancing techniques were used in the hybrid approach, however, these techniques did not consider dealing with various APA parameters simultaneously while also considering their levels of significance in the decision-making process presented in the APA process. Therefore, this research aims to establish a more robust decision-making mechanism for Hybrid Li-Fi networks for APA and NS problems by proposing an integrated Multi Criteria Decision Making (MCDM) solution. The proposed MCDM integrated methodology utilizes Analytic Hierarchy Process (AHP) and ViĊĦekriterijumsko Kompromisno Rangiranje (VIKOR) for factors/criteria weighting and alternatives ranking of AP and NS. All mentioned problems above were addressed using the proposed MCDM mechanism through 4 main phases, including, identification phase, weighting phase, selection phase, and evaluation phase. Different experimentations and simulations were conducted, and the results presented a unique solution which is different and more efficient in comparison with the benchmark works utilized in APA and NS tasks. Moreover, an enhanced LB approach was developed which provides better higher system throughput and increases user fairness for the connected users. In addition, the proposed work aims to reduce the HO rate. The proposed work findings indicate significant improvement that outperforms the benchmark works regarding system throughput, user fairness, HO rate, and switching probability. The proposed approach has been validated and evaluated using two means, that include sensitivity analysis, and MATLAB simulations.