Information Reconciliatioin For Discrete-variable Quantum Key Distribution With LDPC Codes

Cryptography is an important technology in Secure Communications. Compared with the classical key technology depending on the algorithmic complexity, Quantum Key Distribution(QKD) protocol uses physical properties of the Quantum mechanics to ensure an "absolute security" key. Information Reconciliation is an essential part of QKD protocol, which mainly uses the Binary, Cascade and Winnow protocol. As a good channel coding technology, LDPC code perform well for error-correcting and have only one interactive number. This paper is focusing on the application of LDPC codes for Reconciliation.Firstly, the paper simulates the information reconciliation with Binary, Cascade and Winnow protocol. The results show that Cascade performs well at low bit error rates but all of them need too many interactions. Secondly, an alternative promising method employing LDPC codes based on hard decoding algorithm is proposed. The experiment indicates that this strategy extended error-correcting ceiling to 11%. Then, according to the BSC channel model, an optimization algorithm is proposed. From numerical simulation while the performance of our proposed methods in terms of final bit error rate(BER) and disclosed bits is superior to conventional Cascade protocols when QBER is more than 5%. Finally, a method of estimating for source BER based on LDPC codes is proposed, the result of simulation indicates that this method can effectively predict the QBER reducing 1/4 of the disclosed bits.