Web-aplikacija za pohranu, obradu i analizu rezultata mjerenja performansi i mrežnog prometa usluge YouTube

master's thesis

• Title (english): Web Application for Storing, Processing, and Analyzing YouTube Performance Metrics and Network Traffic
• Author: Marina Šolčić
• Mentor: Lea Skorin-Kapov (mentor)
• Committee member: Lea Skorin-Kapov (predsjednik povjerenstva)
• Committee member: Ivana Podnar Žarko (član povjerenstva)
• Committee member: Mirko Sužnjević (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-07-12, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Pružatelji mrežnih usluga uglavnom nemaju uvid u iskustvenu kvalitetu krajnjih korisnika usluge YouTube zbog toga što je sav promet šifriran. Obzirom na te uvjete, pružatelji mrežnih usluga obično ne mogu otkriti KPI-jeve kao što su početno kašnjenje, zaustavljanje videa ili promjene kvalitete iz prometa koji prolazi kroz njihove mreže.
Na Fakultetu elektrotehnike i računarstva izrađeno je mnogo eksperimenata kako bi se razvila rješenja za procjenu iskustvene kvalitete i ključnih indikatora performansi za strujanje videa iz šifriranog prometa. Ti su pokusi napravljeni u kontroliranom ili stvarnom okruženju. Tijekom analize skupa podataka, gdje je skup podataka niz eksperimenata, bilo je mnogo ponavljajućeg rada. Drugi je problem bio u organiziranom pohranjivanju podataka i nedostatku alata za učinkovit pristup, spajanje i filtriranje podataka. Cilj ovog diplomskog rada bio je razviti web aplikaciju koja će riješiti navedene probleme. Vrijeme za analizu skupova podataka bit će značajno smanjeno s razvijenom aplikacijom Auto-A Storage jer će se sva obrada podataka obavljati automatski. Ova aplikacija će pružiti prilagođeno rješenje za pohranu svih mjernih podataka. Neke od glavnih značajki Auto-A Storage-a su spremanje, uređivanje, preuzimanje i filtriranje skupova podataka, kao i preuzimanje generiranih podataka spremnih za strojno učenje. Implementirana je provjera autentikacije i autorizacije tako da nitko ne može pristupiti Auto-A Storage-u bez dozvole administratora aplikacije. Korisnično sučelje Auto-A Storage-a implementirano je u JavaScript biblioteci React, dok je back-end implementiran u Spring Boot-u, u programskom jeziku Java.

Abstract (english)

Network providers for the most part lack insight into YouTube end-users' QoE because of traffic encryption. Given these conditions, network providers usually cannot detect KPIs like initial delay, stalling events or quality switches from the traffic passing through their networks.
At the Faculty of Electrical Engineering and Computing, University of Zagreb, a lot of experiments were performed to collect both application and network data, and develop machine learning-based solutions for estimating video streaming QoE/KPIs from encrypted traffic. Those experiments were performed in a controlled or a real environment. During dataset analysis, where a dataset is a collection of experiments, there was a lot of repetitive work. Another problem was related to storing data in an organized way, and lack of tools for efficient access, merging, and filtering of the data. As a result, the aim of this thesis was to develop a web application that could solve listed problems. The time to analyze datasets will be significantly reduced with developed application Auto-A Storage since all processing of the data will be done automatically.the This application will provide a customized solution for storage for all measurement data. Some of the main features of Auto-A Storage are storing, editing, downloading, and filtering of datasets, as well as downloading generated ML-ready data. It has implemented authentication and authorization so nobody can access to Auto-A Storage without application administrators' permission. The front-end of Auto-A Storage was implemented in React, a JavaScript library, while the back-end was implemented in Spring Boot, a Java framework.

• Keywords: YouTube; web aplikacija; Auto-A Storage; mjerenja; setovi podataka; eksperimenti
• Keywords (english): YouTube; Web Application; Auto-A Storage; Measurements; Datasets; Experiments
• Language: english
• URN:NBN: urn:nbn:hr:168:155181
• Study programme: Title: Information and Communication Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka informacijske i komunikacijske tehnologije (magistar/magistra inženjer/inženjerka informacijske i komunikacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2020-01-07 22:28:46