On the basis of illustrating the Kim's fair non-repudiation protocol,the LSG's protocol is analyzed and the unfairness caused by replay attack is pointed,then a new improvement of Kim's protocol is presented and the fairness and accountability is analyzed by the new Kailar logic.
Zhou-Gollmann protocol is a fair non-repudiation protocol which is widely discussed in recent years.Kim found a flaw of this protocol on time-limited fairness.Aimming at this flaw,Kim gave an improved protocol,but the improved schema highly depended on synchronization of network clock.We analyzed Kim's protocol in detail and show that,without clock synchronization,Kim's improved protocol is also unfair.Then we proposed a new schema to improve it further.Without reducing the efficiency of the original protocol,our new method avoids depending on synchronization of the clock,and keeps fairness and non-repudiation of the protocol.
We deal with two claws of Zhou-Gollmann's fair non-repudiation protocol. Firstly, their protocol divides a message into 2 parts, i.e., a key K and a ciphertext C. Then, C is delivered to the recipient, while K is submitted to TTP (Trusted Third Party). If the originator doesn't submit K to TTP, then the protocol appears to have no dispute between the originator and the recipient. However the protocol depends on his action on whether the originator really submits K to TTP or not. We show that the originator can make the protocol unfair by using his advantageous position, and present how to improve the fairness of the protocol. Secondly, the protocol doesn't provide the message privacy. This means that additional protocols are required to transfer an important message in private. We propose an improved version of the protocol to guarantee the message privacy.