In a traditional (t, n) threshold secret sharing scheme, a secret is divided into n shares and each shareholder keeps one share. Any t or more shareholders can reconstruct the secret if each of them releases its share to the others privately. However, if the communication is over insecure networks, participated shareholders may not be able to verify identities of others and the communicated information may be captured. In these cases, the secret will be obtained by an adversary without a valid share. If an adversary captures messages among shareholders to compute the secret, the attack is passive attack. If an adversary impersonates a legal shareholder to participate in secret reconstruction, the attack is active attack. In order to thwart the two attacks, this paper proposes a novel threshold changeable secret sharing scheme which can realize secure secret reconstruction over insecure networks. In the scheme, legal shareholders use random integers to protect their original shares to resist active attack. Besides, a shareholder is required to keep two shares. It keeps one share private to recover the secret for itself and uses the other share to send to other shareholders. In this way, the scheme can also thwart passive attack.

在传统的（t，  n）门限机密共享方案中，机密分为n股，每个股东保留一股。任何t如果每个人都私下向其他人释放其股份，则更多或更多的股东可以重建秘密。但是，如果通信是通过不安全的网络进行的，则参与的股东可能无法验证其他人的身份，并且所捕获的信息可能会被捕获。在这些情况下，机密者将在没有有效共享的情况下获取秘密。如果对手捕获股东之间的消息以计算机密，则攻击是被动攻击。如果对手冒充合法股东参加秘密重建，则该攻击为主动攻击。为了阻止这两种攻击，本文提出了一种新的阈值可变秘密共享方案，该方案可以在不安全的网络上实现安全的秘密重建。在计划中 合法股东使用随机整数来保护其原始股份，以抵抗主动攻击。此外，股东必须保留两股股份。它将一股私有化，以为自己找回秘密，然后将另一股发送给其他股东。这样，该方案还可以阻止被动攻击。