PSO Algorithm on PAPR Reduction Performance of Asymmetrically Clipped Optical‐OFDM

This research presents a novel hybrid optimization framework combining artificial neural networks (ANN) and particle swarm optimization (PSO) to mitigate the high peak‐to‐average power ratio (PAPR) in asymmetrically clipped optical‐orthogonal frequency division multiplexing (ACO‐OFDM) systems. While conventional PTS effectively reduces PAPR, it suffers from high computational complexity and suboptimal phase factor selection. We propose an ANN‐aided PSO (ANN‐PSO) mechanism where a lightweight feed‐forward neural network dynamically tunes the PSO inertia weight ( ω ) and acceleration coefficients ( c 1 ,​ c 2 ) during the search process, balancing global exploration and local exploitation. Simulation results demonstrate that the proposed ANN‐PSO‐PTS scheme achieves a PAPR reduction of 6.9 dB at a CCDF of 10 −4 for 16‐QAM modulation, outperforming standard PSO and SLM techniques by approximately 1.2 dB. The proposed model demonstrates distinct advantages over established baselines: It provides a 0.3 dB improvement over selective mapping (SLM) and 0.6 dB over tone reservation (TR) without incurring the spectral efficiency penalty associated with reserved tones. Furthermore, the ANN‐adaptive mechanism successfully overcomes the premature convergence observed in standard PSO (which stagnates at ∼8.2 dB), while the optimized swarm search reduces computational runtime by approximately 73% compared to the SLM baseline. The system maintains a robust bit error rate (BER) performance of 10 −5 at reduced SNR levels compared to conventional ACO‐OFDM. These findings confirm the efficacy of metaheuristic‐neural hybrids in delivering robust, low‐complexity optimization for optical wireless communication systems.