隨著網路的進步，網頁上的服務也越來越多，以往的靜態網頁已經演變成動態網頁，你可以和網頁作互動。惡意的使用者具有惡意的瀏覽行為，如洪水攻擊，不斷的浪費資源和網頁主機的頻寬。如今，越來越多的網頁服務已經搬往雲端上面，應用層上的洪水攻擊對雲端的網頁服務並沒有太大的影響效果，但是，雲端的資源也會因為洪水攻擊而造成浪費，增加了計算成本。因此，在此論文中我們提出了植基於語意規則的應用層洪水攻擊之偵測來降低成本。 Along with the progress of the Internet, more and more applications are provided with Web services. The content of the web page has evolved from static into dynamic. People are able to interact with the web page. However, some malicious users have malicious browsing behaviors, such as flooding attacks, to waste the resources and bandwidth of the host of the web page. Nowadays, more and more web services are built on cloud computing. Flooding attacks on the application layer may have no ability to cause denial of service to a Web server on cloud computing, but resources wasted upon attacks still mean unwanted cost on the cloud. Any waste of resources will cause unnecessary cost. Therefore, in this thesis we analyze PHP dynamic pages. According to analysis, we propose a method based on semantic concept to formulate rules to identify malicious browsing behaviors in order to reduce the cost.
