Macro- and microscopic analysis of the internet economy from network measurements

• Autores: Jakub Mikians
• Directores de la Tesis: Pere Barlet Ros (dir. tes.), Nikolaos Laoutaris (dir. tes.), Josep Solé Pareta (dir. tes.)
• Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2016
• Idioma: inglés
• Tribunal Calificador de la Tesis: Jordi Domingo i Pascual (presid.), Josep Lluís Marzo i Lázaro (secret.), Jorge Enrique López de Vergara Méndez (voc.)
• Programa de doctorado: Programa Oficial de Doctorado en Arquitectura de Computadores
• Materias:
  • Matemáticas
    • Ciencia de los ordenadores
      • Informática
• Enlaces
  • Tesis en acceso abierto en: TDX
• Resumen

  • The growth of the Internet impacts multiple areas of the world economy, and it has become a permanent part of the economic landscape both at the macro- and at microeconomic level. On-line traffic and information are currently assets with large business value. Even though commercial Internet has been a part of our lives for more than two decades, its impact on global, and everyday, economy still holds many unknowns. In this work we analyse important macro- and microeconomic aspects of the Internet. First we investigate the characteristics of the interdomain traffic, which is an important part of the macroscopic economy of the Internet. Finally, we investigate the microeconomic phenomena of price discrimination in the Internet. At the macroscopic level, we describe quantitatively the interdomain traffic matrix (ITM), as seen from the perspective of a large research network. The ITM describes the traffic flowing between autonomous systems (AS) in the Internet. It depicts the traffic between the largest Internet business entities, therefore it has an important impact on the Internet economy. In particular, we analyse the sparsity and statistical distribution of the traffic, and observe that the shape of the statistical distribution of the traffic sourced from an AS might be related to congestion within the network. We also investigate the correlations between rows in the ITM. Finally, we propose a novel method to model the interdomain traffic, that stems from first-principles and recognizes the fact that the traffic is a mixture of different Internet applications, and can have regional artifacts. We present and evaluate a tool to generate such matrices from open and available data. Our results show that our first-principles approach is a promising alternative to the existing solutions in this area, which enables the investigation of what-if scenarios and their impact on the Internet economy. At the microscopic level, we investigate the rising phenomena of price discrimination (PD). We find empirical evidences that Internet users can be subject to price and search discrimination. In particular, we present examples of PD on several ecommerce websites and uncover the information vectors facilitating PD. Later we show that crowd-sourcing is a feasible method to help users to infer if they are subject to PD. We also build and evaluate a system that allows any Internet user to examine if she is subject to PD. The system has been deployed and used by multiple users worldwide, and uncovered more examples of PD. The methods presented in the following papers are backed with thorough data analysis and experiments.