Performance modeling of hybrid RF/FSO backhaul systems under severe climate conditions

Abstract Floods and cyclones routinely disable terrestrial fiber and cellular backhaul through physical damage, power loss, and traffic overload, precisely when emergency coordination depends most on connectivity. Free-space optical (FSO) links provide fiber-class capacity without buried infrastructure but are impaired by fog and dense rain-spray, whereas radio-frequency (RF) links remain resilient to optical scattering yet attenuate under heavy rain. This paper proposes a weather-adaptive hybrid RF/FSO backhaul and develops a closed-form, time-resolved availability model driven by a unified meteorological process. The FSO sublink incorporates Beer–Lambert attenuation using the Kim visibility model, Gamma–Gamma scintillation, and Farid–Hranilovic pointing errors; the RF sublink incorporates ITU-R P.838 rain attenuation and Rician fading. Exact outage expressions are derived using Meijer-G and Marcum-Q functions. Selection combining (SC) and maximal-ratio combining (MRC) are analyzed, demonstrating that temporal anticorrelation between fog and rain reduces hybrid outage below the independent-failure product. The framework is evaluated using a 48-h reconstructed trace anchored to official India Meteorological Department (IMD) observations for Severe Cyclonic Storm Michaung, Chennai, December 2023. The hybrid achieves 99.99 % availability, compared with 96.10 % for FSO, 98.07 % for Ka-band satellite, and 99.72 % for RF, exceeding the 99.9 % disaster-grade target. Analytical and Monte-Carlo results agree within 10 −3 . These results establish a mathematically tractable basis for resilient emergency backhaul design under rapidly varying tropical weather conditions encountered.