New approach to quantum key distribution via quantum encryption

Recently, Zhang, Li and Guo (ZLG) suggested a new approach to quantum key distribution by using a shared Bell state which acts as quantum key in order to encode and decode classical information. Subsequently, others extended ZLG protocol to d-dimensional systems and to quantum secret sharing based on reusable GHZ states. However, Gao et al. have shown that if Eve employs a special strategy to attack, these protocols become insecure. Afterwards, they repair ZLG protocol so that their eavesdropping strategy becomes inefficient. In this paper, we investigate the security of ZLG quantum key distribution protocol and show that it is not secure against Eve's attacks and with probability of one half she gets all of the keys without being detected by the two parties. In this eavesdropping strategy, Eve transforms the previously shared Bell state between Alice and Bob to two Bell states among herself and the parties. Moreover, we briefly show that ZLG's repairing by Gao et al's is not efficient against of our attack and Eve can choose an appropriate rotation angle and measurement bases which help her to do eavesdropping. Afterwards, we discuss generalization of ZLG protocol to d-dimensional systems and show that with probability 1/d, Eve gets all of keys. We show that quantum secret sharing based on reusable GHZ states is also not secure and with probability one half, Eve gets all of keys. We repair them by going to higher dimensional shared EPR or GHZ states. Finally, we compare ZLG protocol with ours and show that the ZLG protocol and its extensions are less robust against the channel noise with respect to ours.