臺灣大學: 電機工程學研究所郭斯彥黃敏純Huang, Min-ChunMin-ChunHuang2013-03-272018-07-062013-03-272018-07-062010http://ntur.lib.ntu.edu.tw//handle/246246/254089對於一個給定的惡意程式,我們提供了一個不需要病毒特徵碼並可以即時地回報系統內受感染的檔案。一般而言,雖然惡意程式可以被商業軟體移除,但通常一些相關的惡意元件(例如:幕後主使者(instigator))並未被同時移除乾淨導致惡意程式有再滋生的可能並且可持續地竊取機密資訊或使得我們的系統曝露在公開場合底下而顯得不安全。 此次研究中,我們提供了一個產生感染圖(infection graph)的演算法來關聯起惡意程式及其相關元件;同時,有了感染圖我們除了可以完整地移除單一惡意程式外,亦可基於系統中不同的惡意程式會共用的系統檔案來偵測到其它的惡意程式並移除之。 最後我們做了一個實驗,即將我們的系統所提供的惡意檔案清單和市售防毒軟體的比較。實驗結果顯示,無論是已知或是未知的惡意程式,使用我們的系統可以找到相較於市售軟體還要多的惡意檔案。We provide a real-time system to list all the malicious components for a given malware without the need of any virus definition file. Although now a malware can be detected and removed by commercial tools, however, the related malicious components (called instigator) may not be detected thus malware continuously sacrifice our privacy and expose our system to be insecure. In this study, we provide infection graph generation algorithm to correlate malware and its related malicious component. We can also detect other malwares based on the shared malicious components between malwares. Further, we provide a file list of malicious components and make a comparison with commercial tools. The result of our extensive experiment shows that with our system, we can detect more malicious files than commercial tools for both known and unknown malware.10196360 bytesapplication/pdfen-US惡意程式分析惡意元件偵測感染圖病毒特徵碼系統呼叫程式malware analysismalicious component detectioninfection graphsignaturesystem call惡意程式可疑檔案關聯建立與辨識之即時系統設計與實作Pirus : A Real-Time Framework for Suspicious Entities Correlation and Discrimination for Malware Identificationthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/254089/1/ntu-99-R97921077-1.pdf