Machine Learning–Based Multiparameter Performance Predictions of Ultra‐Dense WDM FSO–FTTx Systems Under Diverse Weather Conditions

Free‐space optical (FSO) communication is vital for modern wireless systems due to its high data rates, energy efficiency, secure transmission, and cost‐effectiveness. However, weather‐induced attenuation, turbulence, and pointing errors affect performance. Recent advancements leverage machine learning (ML) to predict and enhance system reliability under adverse conditions, improving channel awareness and quality of service (QoS). This study has applied modulation schemes, including on‐off keying with nonreturn‐to‐zero (OOK–NRZ), quadrature phase shift keying (QPSK), polarization division multiplexed QPSK (PDM–QPSK), and polarization shift keying (PolSK) with an ultra‐dense wavelength division multiplexed (WDM) FSO‐fiber‐to‐the‐x to modulate the data across clear, fog, rain, and hazy conditions. Furthermore, multiparameter performance predictions of the system under diverse weather conditions have been assessed using ML algorithms such as extreme learning machine (ELM), support vector machine (SVM), and gradient boosting (GB). Mean squared error (MSE) and coefficient of determination ( R 2 ) statistical measures have been employed to measure the robustness of ML. Modulation schemes, atmospheric attenuation, input powers, and FSO link length are fed to the input of ML as features. At the same time, the primary modeling targets are optical signal‐to‐noise ratio (OSNR), quality factor (QF), bit error rate (BER), data rate, and received power. The simulation results have demonstrated that the GB ML achieved the model’s best performance with reduced MSE values for OSNR, QF, BER, data rate, and received power of 2.0203e − 03, 9.3905e − 04, 3.5214e − 12, 1.2527e − 03, and 2.0203e − 03, respectively. Moreover, the study demonstrated exceptional R ‐squared values of 0.9997, 0.9998, 0.9987, 1, and 0.9997, respectively.