🤖 AI Summary
In continuous-variable quantum key distribution (CV-QKD), low throughput and poor real-time performance of low-density parity-check (LDPC) code decoding via the sum-product (belief propagation, BP) algorithm hinder practical deployment. To address this, we propose an optimized BP decoding method incorporating an adaptive early-stopping criterion: posterior probabilities are dynamically monitored during iterations, and convergence-triggered termination eliminates redundant iterations without compromising bit-error-rate (BER) performance. Simulation and experimental results demonstrate that the proposed strategy increases secure key rate throughput by up to 182%, significantly enhancing the real-time processing capability and engineering feasibility of CV-QKD systems. The core innovation lies in the first integration of an adaptive early-stopping mechanism into the BP decoding pipeline for CV-QKD, enabling joint optimization of error-correction performance and decoding speed.
📝 Abstract
We analyse the impact of log a-posteriori early termination on the decoding throughput of reconciliation for continuous-variable quantum key distribution. Increases in decoded secret key rate throughput of up to 182% are reported in both simulations and experiments.