可抵擋入侵的代理簽章系統

Abstract

我們提出一個可抵擋入侵的代理簽章系統。首先，我們設定一角色稱為Home Base，它的用途是與代理簽章者共享代理參數，演進及更新代理簽章者的金鑰。再者，原始簽章者產生一組代理參數，並把代理參數分成兩部分，一份傳給代理簽章者，另一份傳給Home Base。當代理簽章者和Home Base接收到各自的代理參數時，會驗證其有效性。若驗證的結果正確，就接受，否則就拒絕。若代理簽章者要代替原始簽章者簽署一份文件時，他必須先透過Home Base的幫忙去計算出正確的代理秘密金鑰，才可以作簽章的動作。之後，任意第三者都可以驗證其簽章的有效性。此外，我們提出兩個定理證明系統的安全性根基於Strong-RSA problem。第一個定理是證明攻擊者攻破任意個Home Base或代理簽章者，只要不是同時攻破這兩個角色，我們的系統都是安全的。第二個定理是證明攻擊者若同時攻破Home Base和代理簽章者，我們的系統仍可達到前向式安全的特性。

We proposed intrusion-resilient proxy signature schemes, which combine time bounded proxy signature schemes and signer-base intrusion-resilient signatures. First, we set up a home base that shares proxy parameters with the proxy signer, updates and refreshes the proxy signer’s key. Second, the original signer generates proxy parameters and then divides them into two parts. The first part of them is sent to the proxy signer and the second part is sent to the home base. The proxy signer and the home base check the validity of those parameters. If it is correct, they accept it; otherwise, they reject it. Before signing a message m, the proxy signer computes the proxy secret key with the home base’s help. After computing the proxy secret key, the signing and the verification stage are the same as those of time-bounded proxy signature schemes. We give two theorems to prove the security of our scheme by assuming hardness of the Strong-RSA problem.