- 學位論文
植基於使用者匿名性暨身分竊盜之研究
On the Study of User Anonymity and ID-theft
摘要
當使用者想要存取遠端伺服器時，是經由公開通道，所以驗證的機制是必要的。 然而在使用者跟伺服器驗證的過程中所傳的訊息有可能會遭受攔截或竄改， 現今有很多的驗證機制可以去避免遭受這些攻擊。 在商業上有許多的場合中，使用者可能因為個人的隱私的關係而不希望其他人知道使自己所做的商業行為或個人行蹤。 對於在沒匿名的情況下的訊息傳送過程中，如果訊息的內容中含有固定的獨特資訊，則容易被有意圖的人利用此固定獨特資訊來追蹤使用者所做過的商業行為或使用者的行蹤，例如: 姓名， 身分證識別號，信用卡號碼...等等，或利用上述個資衍生出之獨特訊息。 當一位惡意的合法使用者得知另一位使用者的固定獨特資訊，且惡意的使用者未經過此正當的使用者同意，並且以這些固定獨特資訊來做一些像是詐欺或是其它犯罪行為，我們稱之為身分竊盜的攻擊。 因此，在訊息傳送的過程中，必須避免在公開通道上傳送固定的獨特資訊，避免洩漏任何人所做的商業行為或洩漏個人行蹤。 當使用者所傳出訊息的內容是獨特固定訊息，則此獨特固定訊息將可以被用來做字典攻擊，也許可得知使用者的秘密資訊。 所以我們必須讓傳送的獨特固定資訊是動態的，使無法從傳送的訊息中得知獨特固定資訊。 最近匿名性越來越受人們重視， 但是卻沒有一個標準可以得知是否達到匿名，所以在本篇方法中，我們定義了使用者匿名性的分類。 我們將使用者匿名性分為三類: (1) 第一類使用者匿名性: 對所有角色匿名，包含伺服器；(2) 第二類使用者匿名性: 對所有角色匿名，不包含伺服器；(3) 第三類使用者匿名性: 對所有角色匿名，不包含伺服器和其他合法使用者。 另外，要避免身分竊盜攻擊必須滿足此特性: 沒有任何人可以利用使用者與伺服器之間的交換訊息，來代替此使用者存取伺服器，並且在未經過使用者同意下，做一些犯罪的行為。
並列摘要
When a user wants to access a remote server over a public channel, the authentication mechanism is required. However, the messages in the communication of a user to a server may be suffered to interception or tampering. There are many authentication mechanisms can prevent these types of attacks nowadays. In many business situations, the user may hope that no other people can know he/her commercial activities or personal whereabouts. For the situation of message exchange without anonymity and the exchanged message containing a fixed particular information about user ID, a malicious user can easily use the fixed particular information to track the business activities of the user by eavesdropping to the message exchanges. This type of fixed particular information can be: Name, ID number, Credit card number, etc, or information derived from the mentioned data. When a legitimate but malicious user learned that another user's fixed particular information, the malicious user can use this information to do an illegal action, such as frauds or other criminal acts, without the agreement of legitimate user. We call this type of attacks ID-theft attacks. Therefore, in message exchanges, you must avoid to include this type of information. It can prevent the leakage of your commercial activities or personal whereabouts. When user sending a fixed particular message on the public channel, may be easily dictionary attacked to get the user secret information. Thus, we must keep the sending message dynamic as well as possible that cannot learn such a fixed information from sending message. Recently, people pay more and more attention on the user anonymity. Since, there is no mechanism to learn whether does a scheme achieving the user anonymity or not. In our study, we would define a mechanism to give a user anonymity level. We classified user anonymity into three types: (1) The Type 1 User Anonymity: None else including server, can learn user ID. (2) The Type 2 User Anonymity: None else not including server, can learn user ID. (3) The Type 3 User Anonymity: None else not including server and other legal users, can learn user ID. The ID-theft prevention must achieve the property that none else can use a user's exchanged informations to substitute the user to do illegal action, without agreement of the user.