FSO Physical-Layer Security

Power that the receiver and an off-axis eavesdropper collect from a Gaussian beam, their SNRs and the wiretap secrecy capacity for a free-space optical link.

Model: Illustrative

Bob's received power

0.808 dBm

beam radius 0.197 m at Bob

Eve's received power

-18.1 dBm

beam radius 0.197 m at Eve

Bob / Eve power ratio

18.9 dB

optical power

SNR Bob / Eve

65.6 / 27.9 dB

(P / (NEP·√B))²

Secrecy capacity

12.5 Gb/s

12.5 bit/s/Hz × B

Beam divergence

0.395 mrad

full angle 2λ/(πw₀)

Eve's power and the secrecy capacity vs her offset from the beam axis

Each receiver collects the part of the Gaussian beam (radius w(z) = w₀√(1 + (z/z_R)²)) that falls on its aperture, computed exactly, so no receiver gets more than the transmitted power even in the near field. Eve is assumed not to block Bob's beam.

Secrecy capacity of the Gaussian wiretap channel: C_s = B[log₂(1 + SNR_B) − log₂(1 + SNR_E)]⁺ (Leung-Yan-Cheong and Hellman, 1978), with SNR = (P/(NEP·√B))² for receivers limited by the same amplifier noise. This is an illustration of why a narrow beam is hard to tap; an intensity-modulated link is not a Gaussian channel, and the figure is not a security guarantee.

Model and references

Illustrative. A simplified model that shows trends and orders of magnitude. Don't use its numbers for design. About the tiers

Exact Gaussian-beam capture (vacuum, no turbulence; Eve sits before Bob, off his path); SNR (P/(NEP·√B))² with Bob's NEP for Eve too; Gaussian wiretap capacity applied to an IM/DD link. Not a security guarantee.

  1. Leung-Yan-Cheong S. K., Hellman M. E. (1978). The Gaussian wire-tap channel. IEEE Trans. Inf. Theory 24(4), 451–456. Secrecy capacity C_B − C_E. doi:10.1109/TIT.1978.1055917
Report a problem

A wrong value, a formula, a reference, or a page that misbehaves. The report includes this page and its current inputs.