This study presents a comprehensive performance evaluation of the decoy state DL04 quantum secure direct communication (QSDC) protocol over near ground free space horizontal links, with particular emphasis on the combined effects of atmospheric turbulence, aerosol extinction, and background radiation. Unlike previous works that often focus on quantum key distribution or treat environmental factors in isolation, we develop an integrated analytical framework that simultaneously accounts for spatial diffraction of Gaussian beams, Beer–Lambert extinction due to molecular and aerosol scattering, and turbulence induced beam broadening and wandering described by the HV 21/HV Night turbulence models. The turbulent instantaneous transmittance is modelled by a Weibull distribution, from which the average turbulence induced loss is derived. The system performance is quantified by the asymptotic secure capacity under collective attacks and photon number splitting attacks, with single and two photon error rates tightly estimated via a four decoy state protocol. The quantum bit error rate is formulated to include detector dark counts, afterpulsing, intrinsic optical errors, and, crucially, the bidirectional background noise contributions from both the forward (Bob–Alice) and round trip (Bob–Alice–Bob) paths.
Numerical simulations are carried out at 1550 nm wavelength, using typical commercial parameters for both avalanche photodiode (APD) and superconducting nanowire single photon detector (SNSPD). Key quantitative findings are as follows. (1) The receiving aperture and initial beam size exhibit non monotonic optimization: increasing the aperture improves signal collection but also enhances background noise ingress, especially under strong daytime irradiance, leading to an optimum aperture for maximum secure distance; similarly, an optimal initial beam radius exists that balances far field divergence and aperture overfill. (2) SNSPD improves the secure capacity by a factor of 2–3 compared with APD at short to medium ranges, yet the maximum secure communication distance differs marginally between the two detectors. This result indicates that the ultimate range is fundamentally limited by atmospheric channel attenuation rather than detector sensitivity. (3) Turbulence strength and atmospheric visibility are the dominant environmental determinants: below 5 km visibility, extinction overwhelms all other losses and masks day–night differences; above 50 km visibility, further improvement yields diminishing returns as diffraction and turbulence become the residual limiting factors. (4) Daytime background irradiance (~10^-5 W/(m²·sr·nm)) severely degrades both secure capacity and range, necessitating aggressive spatial, spectral, and temporal filtering; in contrast, nighttime irradiance (as low as 10^-8) contributes negligible noise, making nocturnal operation highly favourable for high rate secure communication. (5) Elevation dependence shows that performance saturates above approximately 700 m during daytime and above 350 m at night, suggesting that modest altitude gain can significantly mitigate near ground turbulence.
These results provide a quantitative, physically transparent basis for link budget design, site selection, detector choice, and operational scheduling in practical free space QSDC deployments. The model also reveals the distinct regimes where diffraction, extinction, or turbulence dominates, offering guidance for adaptive system optimisation under varying weather and diurnal conditions.