QoS Technologies and Services in Edge/Fog Computing

Abstract

This thesis investigates Quality of Service (QoS) mechanisms in wireless networks, focusing on the IEEE 802.11e/WMM standard evaluated on a Cisco Aironet testbed. Three router configurations were tested across 162 experimental runs, examining DSCP traffic classes under varying YouTube background load conditions to simulate realistic network environments. A Python-based analysis pipeline was developed to process packet capture data, applying rolling statistical features and unsupervised clustering methods including KMeans and DBSCAN to characterize traffic behavior. The results demonstrate that WMM-based QoS configurations significantly influence latency, jitter, and throughput across upload and download directions, particularly under high background load. Configuration tuning proved effective in prioritizing time-sensitive traffic classes, highlighting the practical value of QoS enforcement at the access point level. The findings contribute empirical evidence toward optimizing wireless QoS deployments in edge and fog computing environments.