| With the development of science and technology,the transmission of information between the communicating parties becomes very easy and convenient.At the same time,information security attracts a great deal of attentions and has become very important for modern organizations.Based on the fundamental principles of the quantum physics(including the uncertainty principle and the quantum no-cloning principle,etc.),quantum key distribution(QKD)enables two remote and legal parties(referred to as Alice and Bob)who are linked by a quantum channel and a classical authenticated public channel to share an unconditionally secure key,which is unknown completely to the potential eavesdropper(Eve).A secure communication can be achieved between two legitimate parties when combine the quantum key distribution with the one-time-pad encryption.Since Charles H.Bennett and Gilles Brassard proposed the first QKD protocol in 1984(BB84 protocol),QKD has attracted great attentions.The past three decades have witnessed rapid progress of QKD both in theory and experiment.Continuous variable quantum key distribution(CVQKD)requires only simple measurement equipment and has good compatibility with the existing passive optical communication networks.Furthermore,CVQKD can achieve a high secure key rate at a relatively short distance.Information reconciliation is a critical procedure in QKD system,which directly affects the final secure key rate.In this thesis,the information reconciliation for CVQKD is investigated.The main contents and innovations of this thesis:1.We propose a designing scheme of irregular LDPC codes with large block length.Three different construction methods of LDPC codes are investigated.It is shown that the random construction method performs superiorly at low rate codes,while quasi-cyclic extension is better for high rate codes.Good degree distributions with threshold close to capacity can be obtained via density evolution and differential evolution algorithm.With above developed techniques,high error-correcting performance irregular LDPC codes are constructed.2.For the signal-to-noise ratios(SNRs)above 1,we achieved highefficiency information reconciliation for CVQKD based on the slice reconciliation protocol with efficiency above 95% and a FER below 24%.The slice reconciliation scheme is divided into two parts: quantization of Gaussiandistributed variables and error correction.By using a 32-equal-interval quantizer,the quantization efficiency reaches above 99.4%.The error correction process is the main factor that influences the reconciliation efficiency.Using the above high performance irregular LDPC codes,good error correction can be achieved.3.We propose a rate-adaptive protocol for continuous variable key reconciliation.In contrast to the fixed code rate information reconciliation,this protocol can cover a relatively broad range of signal-to-noise ratio and achieve higher reconciliation efficiency4.Multidimensional reconciliation for CVQKD is studied in detail.For SNR of 0.4,we realize multidimensional reconciliation with efficiency 92.21% using irregular LDPC codes. |
