Defesa de Dissertação de Mestrado Nº 1.173: "Energy-Cost aware vertical handover decision algorithm based on probabilistic time estimation model for video streaming sessions"
O aluno Sálvio Freire vai defender seu trabalho no dia 10 de agosto, às 15h, no Anfiteatro do CIn
Início: 10/08/2012 às 15:00
Término: 10/08/2012 às 17:00 Local: Anfiteatro do CIn
Pós-Graduação em Ciência da Computação – UFPE
Defesa de Dissertação de Mestrado Nº 1.173
Aluno: Sálvio de Barros Freire
Orientador: Prof. Paulo Roberto Freire Cunha
Título: ENERGY-COST AWARE VERTICAL HANDOVER DECISION ALGORITHM BASED ON PROBABILISTIC TIME ESTIMATION MODEL FOR VIDEO STREAMING SESSIONS
Hora/Local: 15:00 – Anfiteatro
Prof. Kelvin Lopes Dias ( CIn / UFC)
Prof. José Neuman de Souza( Departamento de Computação / UFC)
Prof. Paulo Roberto Freire Cunha( CIn / UFC)
Currently, there has been an increasingly penetration of several wireless access networks/technologies, such as WiMAX, Wi-Fi, UMTS/LTE, along with the popularity of mobile multimedia applications and new services available in mobile terminals. In these heterogeneous scenarios, it is expected that mobile users can take advantage of accessing the best wireless opportunities, e.g., in terms of monetary cost and Quality of Service (QoS) even while using energy-constrained devices. Thus, the energy consumption issue is a key factor that should be addressed in order to support QoS for mobile applications, especially on video streaming sessions.
At the same time, the necessity of optimization of energy consumption raises the cost issue, since these networks demands high charges to the mobile user. Monetary cost of the access is an important factor for avoiding churning and even gain of more carriers. In a heterogeneous environment composed by WiMAX and Wi-Fi, the connectivity opportunities should take cost and battery lifetime, since these are opposite parameters for these two networks. In order to address this issue, on this work is introduced a vertical handover decision algorithm which aims to optimize the battery device and at the same time balance the monetary cost, through a probabilistic model for estimation of the remaining application time, which is retrieved from a historical database of previous sessions.
Simulations with NS-2 were performed with the finality to compare the proposed scheme with methods where the cheapest network is selected first due your lower cost, and statistical analysis demonstrated that on the proposed algorithm is possible to approximate the monetary cost to the traditional scheme on scenarios where battery level is under critical level, and also to optimize at least 7% your energy consumption.
Palavras-chave: Energy, Monetary Cost, Vertical Handover Decision, Time Estimation, Performance Evaluation, Video