A distributed threshold signature algorithm on canon of probability

In a nonsecurity network, threshold signature is used for distributing keys for KDC (key distribution center), which can improve the security of KDC and avoid single-point invalidation. But the traditional threshold signature does not take into account the actual environments of networks; it simply assumes that every server composing KDC has an equal status, but it is not the truth. This paper tries to correct this fault and provides a new way to configure those servers. The core idea is to add a parameter probability weight to every server. It can reflect the security rank and capability (also some other factors) of the corresponding server. This new transaction fits better in complex network environments; it also improves the capability of detecting intrusion from another perspective.