ゼミナール発表

日時: 9月29日(火)1限 (09:20-10:50)


会場: L1

司会: 油谷 暁
田川 真樹 1451066: M, 2回目発表 山口 英,岡田 実,門林 雄基,妙中 雄三
title: Network capacity expansion methods based on efficient channel utilization for multi-channel wireless mesh network
abstract: This study implements an architecture of traffic management on multi-channel wireless backbone network (WBN). The majority of existing studies on multi-channel WBN focus on routing and channel assignment with a limited number of WLAN interfaces in each AP due to hardware specification. On the other hand, there are many vacant channels in the 5 GHz band, which means that wireless resources are not used effectively. Therefore, the aggregation method of all available channels and the traffic management method that fully utilizes the channels are essential. Therefore, we proposed techniques to build an OpenFlow-based multi-channel wireless backbone network and then implement three fundamental channel utilization methods which are based on the traffic volume/airtime. However, the calculation of traffic volume/airtime could be incorrect due to packet loss. Accordingly, these flow allocation methods lead to inefficient flow allocation. In this presentation, we present an on-going channel utilization method that is for packet loss reduction. Furthermore, we prove that the proposed method can reduce packet loss even in complex traffic situations.
language of the presentation: Japanese
 
村上 和有基 1451109: M, 2回目発表 山口 英,藤川 和利,門林 雄基
title: The proposal of detection method for drive-by-download attack based on iframe usage
abstract: Cyber attacks have become a significant threat on the Internet. Drive-by-downloads are a type of these attacks. They use manipulated web pages in order to lead victims to malicious servers such as malware distribution servers. Users who simply access the manipulated site are affected by these attacks. There are some redirection methods to lead victims from manipulated web pages to malicious page. We focus on the iframe usage. We checked iframe usage on benign sites and malicious sites. In this presentation, we propose the detection method for drive-by-downloads based on iframe usage.
language of the presentation: Japanese
 
菅沼 孝二 1451059: M, 2回目発表 萩田 紀博,笠原 正治,浮田 宗伯
title: Prediction of Temperature Distribution by Online Learning Algorithm for Green Data Centers
abstract: Temperature distribution inside a datacenter affects overall power consumption of its facility. Therefore, it enables efficient control of the equipment which is for reducing power consumption of whole datacenter, by predicting temperature distribution inside a datacenter based on information of parameters for equipment settings and data from equipment, censors, etc. In this paper, we propose a method to predict temperature distribution inside a datacenter by applying machine learning with the past observation data inside. In this paper we evaluate prediction of temperature distribution inside a datacenter by the proposed method with the observed data acquired from the datacenter under our research group operation.
language of the presentation: Japanese