Caracterização do comportamento de usuários e precificação de tráfego de internet de banda larga

• Ano de defesa: 2008
• Autor(a) principal: Humberto Torres Marques Neto
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: precificação de tráfego; Sistemas de comunicação em banda larga Aspectos econômicos; Computação; Provedores de serviços da Internet Aspectos econômicos
• Link de acesso: http://hdl.handle.net/1843/RVMR-7K6PHT

Broadband user behavior characterization can lead to better understanding and identifying different classes of users, defined by user request patterns. This characterization could help broadband carriers capacity management and planning efforts. Different service request patterns require different service prices. Nevertheless, the main pricing scheme used by broadband ISPs (Internet Service Providers) is the flat rate pricing with limited peak bandwidth, which does not consider different user behavior patterns. This scheme does not guarantee fairness because it encourages resource wasting and, as a consequence, light users end up subsidizing heavy users. A fair pricing scheme, which fosters balanced resource consumption, could save bandwidth and user's money. Saving resources allows the increase of subscriptions, which could improve ISP profit.This work propose a Internet broadband pricing scheme based on user behavior during the day (BPS -- Broadband Pricing Scheme), which is fair to users and profitable for ISP. The reason to create this scheme is based on the results of broadband user behavior characterization with a proposed methodology and real data from an ISP. The validation and the comparison of proposed scheme with other showed in literature are done through the simulation with the same real data used in broadband user behavior characterization. Started from the characterization with real data, this work shows that BPS fosters fair Internet use and optimizes ISP resources re-distributing the workload during hours of day. The simulation of the proposed scheme and others from the literature shows that BPS is a fairner scheme than others, mainly, because utility function used to calculate user benefits is greater or equal zero in almost hours of day. Furthermore, simulation results shows that bandwidth saving happens in almost 60% of all hours of the day, which is the best result among simulated schemes. The usage of a fair pricing scheme could create benefits to both users and to ISPs.