Optimization Of Error Reconciliation In Discrete Varible Quantum Key Distribution

With the development of quantum computing, the classic cryptography based on computational complexity is now seriously threatened by the strong parallel computing ability of quantum computation, and one-time pad encryption algorithm proved to be absolutely secure can not be widely used because of the high demand for its key. Emergence of quantum key distribution technology enables high-speed, high-security key generation and distribution as possible, but it also provides the conditions for the one-time pad encryption widely used.Error reconciliation module is key module in QKD system, study and optimization of error reconciliation to improve the overall performance of QKD system, promote QKD technology has important significance.Performance of error reconciliation mainly reflected in negotiation efficiency and negotiation rate two aspects, negotiation efficiency is a measure of the ratio of the amount of disclosing information in the negotiation process; negotiation rate is a measure of the amount of handling information per unit of time in the negotiation process. In this paper, optimizes the negotiation efficiency or negotiation rate for several typical error reconciliation algorithm. Details as follows:Propose a negotiation efficiency optimization algorithm for Winnow error reconciliation algorithm. Through the analysis of the original Winnow algorithm, find out that negotiation efficiency of Winnow algorithm is closely related to the length of the block. To determine the optimal block length, we determine the relationship between the the best single round block length and the bit error rate through the emulation mode; determine the error rate estimation scheme through theoretical analysis; finally, applying the idea of gready, proposed an adaptive block length optimization program realize the optimization of Winnow algorithm in negotiation efficiency. The simulation results show that the negotiation efficiency of Winnow optimization algorithm mentioned in this paper is superior to the original Winnow algorithm and its typical improved algorithms.Achieve a negotiation rate improvement program for Cascade error reconciliation algorithm. Through the analysis of the original Cascade algorithm, find out that the number of interations directly affect the negotiation rate of Cascade alogritm. In order to reduce the number of interations, according to the characteristic that the respective blocks and sub-blocks are parallel processing, we design and implement a parity block transfer programs through employing the Open MP programming model, while improving the backtracking strategy and parity calculations. The simulation results show that the proposed optimization can significantly improve the negotiation rate of Cascade algorithm.Achieve two negotiation rate improvement program for LDPC error reconciliation algorithm. Through the analysis of the original LDPC algorithm, determine the optimization direction of the algorithm. First, we determine the parity check matrix, best extended sub, normalization factor and the maximum numbers of iterations by the method of simulation contrast. Secondly, through the study of decoding implementation manner and storing manner of check matrix, determine accelerated program employing GPU technology for implementation manner of check node and variable node seperate updatation, and give employing programming model of the accelerating program and the storage scheme of corresponding check matrix; determine accelerated program employing X86 AVX instruction set for the implementation manner of combining hard decision information and check node information updatation, and give the storage scheme of corresponding check matrix. The simulation results show that the proposed optimization can significantly improve the negotiation rate of LDPC algorithm.