Researches On Noise Reduction Mechanism In Quantum Signals Co-Propagating With Classical Signals

Quantum Key Distribution (QKD) technology, based on the Heisenberg uncertainty principle, quantum measurement collapse principle and quantum no-cloning theorem, can achieve unconditionally secure communication. At present, great progress has been achieved on single-photon-based point-to-point optical fiber QKD system. To extensively deploy QKD system, how to multiplex ultra-weak quantum signals with high-power classical optical signals in the same fiber remains a critical issue to be solved. In this paper, we paid much attention to the researches on noise reduction mechanism in quantum signals co-propagating with classical signals.A user-based channel-interleaving Wavelength Division Multiplexing(WDM) scheme with unequal frequency spacing was proposed. In regard to the interference problem when ultra-weak quantum signals co-propagating with classical optical signals, this paper adequately studied the influence of four-wave mixing (FWM) and Raman scattering on quantum signals and quantum bit error rate of the system, and determined FWM as a dominant position in all of the noise. Then we put forward a user-based channel-inter leaving WDM scheme with unequal frequency spacing to effectively avoid the nonlinear effect and crosstalk between channels on quantum signal in fiber transmission.Simulations and verifications through self-developed FWM effect simulation software were implemented. We utilized Graphical User Interface (GUI) in MATLAB to edit an FWM effect simulation software,realizing the quick settings of the simulation parameters and convenient reading and analysis of the simulation results. Meanwhile verified the feasibility of the user-based channel-interleaving WDM scheme with unequal frequency spacing.An algorithm based on MapReduce, aiming at solving the computing problems of the unequal frequency spacing in the wavelength allocation scheme proposed was presented. This paper analyzed that the computing problems of the unequal frequency spacing was an NP-complete Integer Linear Programming (ILP) at first, and solved it when channel scales were small. Secondly we came up with an algorithm based on MapReduce to settle the insufficient of the ILP problems in time complexity. Finally the MapReduce algorithm was verified more effective by simulation when channel scales increased.