Modeling and analysis of quality of service (QOS) in heterogeneous wireless network

Abstract

Heterogeneous Wireless Network (HWN) is the next-generation wireless communication system platform where Internet services will be offered and delivered through multiple technologies like WiFi, WiMAX, and GSM. Many Internet applications require high bandwidth and therefore, bandwidth aggregation has been considered as one of the effective solutions for high bandwidth required applications in HWN. There are few issues have been raised for bandwidth allocation in HWN including the selection of Radio Access Technology (RAT), bandwidth allocation among these RATs, and packet reordering occurrence leading to quality degradation in HWN. This research aims to model and analyze the RAT section methods, resource allocation policy, and packet reordering pattern. Three models have been followed with the numerical and simulated results to achieve better performances of the proposed models. A multi-criteria selection-based method, Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) has been applied to select the most suitable RAT in HWN. The TOPSIS method has been selected due to its sensitivity to the application requirements. The method has been coded in Matlab and numerical models have been analyzed. The TOPSIS method achieved to select the most appropriate RATs based on the types of services like voice calls, file downloading and video streaming according to their priority settings. Resource allocation has been modeled based on Shapley Value, a game-theoretic approach. The amount of bandwidth has been calculated to transmit 600Kbps data over the three networks where 225Kbps, 141Kbps, and 233.33Kbps have been transmitted through, WiFi, Cellular, and WiMAX networks respectively. Packet reordering is a common phenomenon in HWN that has been investigated briefly to analyze its metrics and pattern. An algorithm has been proposed to measure the associated delay and reordering entropy caused due to packet reordering in HWN. It has been compared to the Earlies Delivery Path First (EDPF) method both for CBR and VBR applications on the NS3 simulator. A video clip for 1800 seconds has been transmitted from sender to receiver to analyze the performance in HWN. Based on the simulation results, the proposed method shows around 40% improvement over the EDPF method both for CBR and VBR applications in terms of packet reordering delay. Moreover, the proposed method has shown a significant achievement for the reordering entropy over the EDPF method for both applications.