一個以基於雲端服務之網路行為分析系統

Abstract

在網路環境中充滿著各式各樣的分散式網路服務與大量資料，因此當網路上增加新的資訊時幫助 企業或使用者在服務導向雲端運算環境中動態且即時取得新資訊，是非常受到重視的。其中，在網路 資訊擷取週期的部分，如果擷取頻率過高，將造成網路負擔和成本的花費；反之，如果擷取頻率過低， 則會使得網路資訊延遲。因此最佳網路資訊擷取週期將是重要的研究議題。然而，在目前現有的資訊 系統中卻缺乏一套有效的網路行為分析模式，以提供基於雲端服務的最佳網路資訊擷取週期推薦。 因此本計畫提出整合網路服務技術與網路行為分析模式的研究，導入雲端運算環境中，搭配語意 網、網路行為分析模式等方法，完成基於雲端服務的網路行為分析系統。研究內容主要包括：(1) 建 構網路行為分析模式、(2) 分析最佳網路資訊擷取週期、(3) 建構服務導向架構之資料傳輸和分析。 本計畫將發展與實作出此方法架構與系統，實際在一個有即時資料變動的領域測試實用，並進行方法 的評估。

It will be full of distributed network services and a large quantity of data in network environment. It is important to dynamically retrieve the real-time network information when the information is published in network. In the determination of Optimal Cycle for Network Information Retrieval (OCNIR), the network loading and the cost of information retrieval are higher when the retrieval frequency is higher. If the retrieval frequency is too low, the delay of information retrieval is long. For this reasoning, the OCNIR is an important issue. However, the OCNIR for cloud computing service has not been investigated. In this research, we provide the integration of web service techniques and Network Behavior Analysis Model (NBAM). We implement Service-Oriented Architecture (SOA) environment with semantic web and NBAM to accomplish a Network Behavior Analysis System (NBAS) for cloud computing service. The major contents of this research are as followed: (1) the establishment of NBAM, (2) the determination of the optimal cycle for network information retrieval, and (3) the establishment of data transmission and analysis based on SOA. We develop the approach and implement the NBAS for individuals or enterprises to find out NBAM and OCNIR, test this approach in a domain with real-time data alteration, and then evaluate this approach.