WDM-PON Free Space Optical (FSO) System Utilizing LDPC Decoding for Enhanced Cellular C-RAN Fronthaul Networks

Abstract

Modern cellular systems rely on high-capacity and low-latency optical networks to meet ever-increasing data demands. Centralized Radio Access Network (C-RAN) architectures offer a cost-effective approach for deploying mobile infrastructures. In this work, we propose a flexible and cost-efficient fronthaul topology that combines Wavelength Division Multiplexing (WDM) passive optical networks (PONs) with free-space optical (FSO) links. To enhance overall system performance, we introduce Low-Density Parity Check (LDPC) decoding, which provides robust error-correction capabilities against atmospheric turbulence and noise. Our system transmits 20 Gbps, 16-QAM intensity-modulated orthogonal frequency-division multiplexing (OFDM) signals, achieving a substantial reduction in bit error rate (BER). Numerical results show that the proposed WDM-PON-FSO architecture, augmented with LDPC decoding, maintains reliable transmission over 2 km under strong turbulence conditions.

1. Introduction

High-capacity, low-latency communication remains a pivotal requirement for next-generation cellular networks, particularly as Centralized Radio Access Network (C-RAN) architectures rapidly evolve to support emerging beyond-5G services [1,2]. Wavelength Division Multiplexing–Passive Optical Network (WDM-PON) systems offer a key solution for high-throughput and cost-effective fiber deployments, thanks to the scalability and dedicated bandwidth afforded by wavelength routing [3,4,5]. Yet, purely fiber-based fronthaul may be impractical in geographically constrained areas or during urgent deployments. Consequently, Free-Space Optical (FSO) links have emerged as a complementary or alternative approach, utilizing unlicensed optical bands to provide rapid installation and substantial bandwidth potential [6]. Despite its advantages, FSO transmission is highly sensitive to atmospheric turbulence—characterized by random fluctuations in air refractive indices [7]—as well as losses due to fog, rain, and pointing errors [8,9].

To combat these impairments, advanced forward error correction (FEC) strategies are indispensable [10]. Low-Density Parity-Check (LDPC) codes, in particular, have garnered significant attention for their near-capacity performance and flexible rate adaptation, making them well suited for hybrid optical–wireless architectures [11].

Unlike traditional forward error correction (FEC) techniques, such as convolutional or Reed–Solomon codes, which primarily correct random bit errors, LDPC codes excel in correcting bursty errors that occur due to atmospheric turbulence in FSO links. FSO channels suffer from intensity fluctuations caused by refractive index variations, leading to deep fading events that create clusters of errors in received data. LDPC’s iterative decoding mechanism mitigates these error bursts by leveraging soft decision decoding and message-passing algorithms to reconstruct lost information. Moreover, LDPC’s structured parity-check matrices allow for flexible code rate adaptation, making them highly effective under varying turbulence conditions, where signal-to-noise ratios fluctuate unpredictably.

Recent studies have demonstrated novel LDPC variants (e.g., nonbinary LDPC, protograph-based LDPC, and generalized LDPC) that substantially improve bit error rate (BER) under strong turbulence or log-normal/Gamma–Gamma channels [12,13,14,15].

Accordingly, merging WDM-PON with FSO and LDPC decoding holds substantial promise for next-generation C-RAN fronthaul. Previous studies on hybrid WDM-PON–FSO systems have demonstrated significant gains in capacity, often achieved through multi-wavelength designs or advanced modulation and coding techniques [16,17,18,19]. However, many existing solutions still face critical challenges that hinder widespread deployment. This paper proposes a WDM-PON Free-Space Optical (FSO) system utilizing LDPC decoding and orthogonal frequency-division multiplexing (OFDM) modulation to enhance the throughput, reliability, and overall robustness of C-RAN fronthaul networks. Specifically, we:

Present a system architecture that integrates multi-wavelength passive optical distribution with atmospheric wireless access, leveraging OFDM to improve spectral efficiency and resilience against channel impairments in 5G/6G deployments.

Implement advanced LDPC codes optimized for strong turbulence conditions, achieving significant BER improvements by mitigating the impact of fading and signal distortions inherent in FSO links.

Conduct comprehensive simulations and theoretical analyses to evaluate system performance under severe turbulence conditions.

By harnessing the combined strengths of WDM-PON, OFDM, FSO, and LDPC-coded optical transmission, our study addresses key performance and cost challenges, paving the way for next-generation C-RAN solutions that scale efficiently under diverse environmental conditions and traffic demands.

The remainder of this paper is organized as follows. Section 2 discusses existing methods from the literature, highlighting the limitations our approach addresses. Section 3 presents the proposed system architecture, and Section 4 details the simulation setup and key performance findings. Finally, Section 5 concludes the work and outlines future research directions.

2. Related Work

Wavelength Division Multiplexing–Passive Optical Network (WDM-PON) technologies have evolved into core solutions for high-bandwidth broadband access. Early approaches focused on point-to-multipoint architectures with dedicated wavelengths per user [3], while subsequent innovations introduced a range of cost-saving measures, e.g., reflective semiconductor optical amplifiers (RSOAs) [4] and advanced arrayed waveguide gratings [20]. Researchers have also explored integrating next-generation standards (e.g., coherent 100 G) to push capacity further [21,22,23]. Meanwhile, multi-wavelength data comb generation [24] and dynamic resource allocation [25,26] underscore WDM-PON’s adaptability in heterogeneous networks.

FSO channels, on the other hand, provide attractive “last-mile” or “fronthaul” solutions due to their high data rates and license-free spectrum [7]. However, atmospheric turbulence remains a dominant impairment, often modeled by log-normal or Gamma–Gamma fading [8,14,27]. Environmental factors (fog, rain, misalignment) can trigger large fluctuations in received signal intensity [16], thus degrading link availability. To bolster reliability, hybrid architectures have emerged: pairing FSO with millimeter-wave or RF backups [28], employing cooperative relay networks [15], or adopting advanced automatic repeat requests [29] and prediction techniques [30].

Recent work has consistently shown that Low-Density Parity-Check (LDPC) codes are promising for bridging performance gaps in atmospheric optical channels [27]. Owing to flexible rate adaptation and near-capacity performance, LDPC codes can significantly reduce BER, even in strong turbulence or multi-hop FSO links [31,32,33]. Multiple decoding algorithms, such as weighted bit flipping, min-sum updates, or protograph extension, have been proposed to handle large block sizes and high data rates [27,34].

Moving beyond discrete deployments, a rising body of literature integrates FSO links directly into WDM-PON backbones to harness high capacity and flexible coverage. Some studies focus on resilience (e.g., fiber-FSO path redundancy or ring-based self-healing topologies) [18,35,36,37], while others target improved spectral efficiency and modulation (e.g., OFDM, quadrature amplitude modulation) [8,17,38].

Although a handful of proposals integrate LDPC-coded designs with WDM-PON-FSO architectures, the majority either overlook strong atmospheric fading or rely on baseline coding that fails to fully exploit the iterative decoding gains of LDPC algorithms [29,39,40]. Similarly, attempts to unify high-speed, multi-wavelength PON distribution with robust FSO wireless links for C-RAN remain relatively sparse [22].

Notwithstanding the abundant research on (1) WDM-PON for broadband access, (2) FSO for fast, flexible wireless links, and (3) LDPC coding for improved optical communications, a coherent approach that addresses harsh turbulence, ensures robust error correction, and seamlessly aligns with C-RAN’s capacity and latency requirements is still lacking. In particular:

Many studies do not fully incorporate advanced LDPC decoders optimized for strong turbulence channels, especially at high line rates.

System-level demonstrations of WDM-PON–FSO fronthaul, optimized for 5G and beyond, remain limited.

While some works propose rate-compatible or probabilistic shaping schemes [27,41,42,43,44], few solutions explicitly combine them with hybrid FSO topologies for cellular fronthaul.

By systematically integrating these components, our research aims to fill the existing gaps, offering a WDM-PON FSO system utilizing LDPC decoding that meets the stringent demands of cellular C-RAN fronthaul. In what follows, we detail the proposed architecture, describe the LDPC-based error correction scheme, and provide comprehensive results illustrating how this approach achieves enhanced reliability, even under pronounced atmospheric turbulence.

3. System Architecture

The proposed WDM-PON-FSO system is designed to provide seamless high-speed connectivity. The system architecture shown in Figure 1 consists of a centralized baseband unit (BBU), an optical line terminal (OLT), multiple optical network units (ONUs), an LDPC encoder and decoder, and a free-space optical (FSO) link that extends connectivity to remote locations.

The centralized baseband unit (BBU) serves as the primary control node, managing both data transmission and reception tasks within the network. To facilitate WDM-PON operations and make efficient use of available bandwidth, it employs a multi-wavelength comb source that generates multiple optical carriers. Moreover, it utilizes digital signal processing (DSP) techniques, most notably LDPC coding and decoding, to minimize errors and maintain high-quality signal performance.

Positioned at the interface with the core network, the Optical Line Terminal (OLT) directs optical signals toward Optical Network Units (ONUs) through a WDM-PON infrastructure. It relies on arrayed waveguide gratings (AWGs) to handle multiplexing and demultiplexing processes, ensuring precise separation of individual channels.

At each subscriber location, an Optical Network Unit (ONU) is responsible for extracting data transmitted by the WDM-PON-FSO configuration. These ONUs also include LDPC-based error correction modules, which improve signal fidelity prior to further data processing.

Optical link turbulence, whether moderate or severe, can be quantified via the Rytov variance. Concurrently, scintillation induces fluctuations in signal intensity, ultimately undermining system performance. To mitigate these effects, advanced Low-Density Parity-Check (LDPC) codes are employed, thereby enhancing the system’s resilience in the face of noise and atmospheric perturbations.

The LDPC encoder adds redundancy during transmission by using a generator matrix to convert a binary input sequence of length k into a binary output sequence of length n. The first $n-k$ bits in the output represent parity information derived from the encoding procedure, while the remaining k bits map directly to the original input sequence. A sparse parity-check matrix H, comprising $n-k$ rows and n columns, underpins effective error correction by satisfying the condition:

$H·c^T=0,$

where c denotes the encoded codeword vector.

At the receiver, the LDPC decoder employs the belief propagation (sum-product) algorithm to iteratively recover the transmitted data from the received noisy version. The decoder uses soft-decision decoding to estimate the original binary sequence and applies the log-likelihood ratio (LLR) calculation:

$\mathrm{LLR}\left(y_i\right)=log\left[{\displaystyle \frac{P(y_i\mid c_i=0)}{P(y_i\mid c_i=1)}}\right],$

where $P(y_i\mid c_i)$ represents the conditional probability of receiving $y_i$ given that the transmitted bit was $c_i$. The decoding process iteratively updates probability estimates, ensuring minimal latency while maintaining high accuracy.

The proposed architecture combines a WDM-PON backbone with a free-space optical (FSO) link, using LDPC coding and OFDM modulation to ensure high reliability. On the transmitter side, a central baseband unit (BBU) hosts a multi-wavelength source (such as a comb laser) feeding an arrayed waveguide grating (AWG) for WDM multiplexing of numerous 1550 nm-band channels. Each optical carrier is intensity-modulated with a high-speed OFDM signal carrying the data. The data stream first passes through an LDPC encoder, which adds redundant parity bits for error correction. Next, the encoded bits are mapped to QAM symbols (16-QAM in this design) and distributed across many subcarriers by an OFDM modulator. Each WDM channel uses a non-return-to-zero (NRZ) pseudo-random bit sequence (PRBS) of length $2^{23}-1$ at a data rate of 40 Gbps. In the OFDM modulator, 512 subcarriers are used along with a 2048-point Fast Fourier Transform (FFT). Consequently, the data rate becomes ($512/1024$) $\times 40=20$ Gbps. To prevent intersymbol interference (ISI), a cyclic prefix of 100 samples is added after the Inverse FFT (IFFT) process. The proposed system supports $N=32$ wavelengths over a WDM FSO link ($32\times 20$ Gbps), covering frequencies between 193.1 THz and 194.65 THz (i.e., 1552.52 nm to 1540.16 nm). This yields a total capacity of 640 Gbps. Prior to transmission, pulse shaping is performed with cosine roll-off filters to further minimize ISI.

The generated QAM-OFDM electrical signal then drives a Mach–Zehnder Modulator (MZM) that impresses it onto a laser beam for each WDM channel. All 32 wavelength channels (each 20 Gb/s) are optically combined, boosted by a gain-flattened EDFA (optical amplifier) to about 10 dBm launch power, and sent into the FSO channel. The FSO transmitter uses a collimating telescope or lens ($\approx 5$ cm aperture) to direct the multi-wavelength optical beam toward the remote receiver through the air. At the receiver end, a collecting telescope ($\approx 20$ cm aperture) gathers the incoming optical beam. The combined WDM signal is then demultiplexed by a DeMUX/AWG, where each wavelength is routed to its corresponding receiver. At the receiver, a PIN photodiode converts the received optical signal into an electrical RF signal, which is amplified by a low-noise amplifier to boost the weak photocurrent. In each channel’s receiver, the electrical signal is fed to an OFDM demodulator, which performs FFT to recover the subcarriers and then de-maps the QAM symbols. These symbols are mapped back to bits, and finally, the stream passes through an LDPC decoder (using iterative belief-propagation) that corrects errors caused by noise or fading. Key hardware components include the optical transmit/receive lenses (antennas) for beam launching and collection, the EDFA, which counteracts propagation loss by amplifying the optical signal before free-space transmission, and the electrical amplifiers at the receiver to strengthen the detected signal. Each component interconnects in a pipeline: the LDPC-coded, OFDM-modulated signal is carried on an optical carrier (via MZM and laser) over the FSO link, and the reverse process (photodetection, OFDM, and LDPC decoding) retrieves the original data at the far end. This design leverages the enormous WDM-PON capacity and extends it wirelessly via FSO, while LDPC and OFDM ensure the link meets the high reliability and throughput needed for C-RAN fronthaul.

FSO Channel Modeling

Atmospheric attenuation and fluctuations in signal amplitude and phase—often referred to as intensity scintillation—play a major role in free-space optical (FSO) communication. Multiple statistical models, such as the Log-normal, K, Negative exponential, Gamma–Gamma, and Log-normal Rician distributions, are utilized to describe optical turbulence [45]. Among these, the Gamma–Gamma model is extensively used due to its effectiveness in capturing a wide range of turbulence effects and related attenuation [6].

The FSO link in the proposed system is modeled using the Gamma–Gamma distribution, which accurately represents the effects of atmospheric turbulence on signal propagation. This distribution accounts for intensity fluctuations due to refractive index variations, leading to signal fading and degradation in free-space optical links. The normalized light intensity I is characterized by the probability density function (PDF) presented in [46]:

(1)$P\left(I\right)={\displaystyle \frac{2{\left(\alpha \beta \right)}^{\frac{\alpha +\beta }{2}}}{\mathrm{\Gamma }\left(\alpha \right)\mathrm{\Gamma }\left(\beta \right)}}{I}^{\frac{\alpha +\beta }{2}-1}{K}_{\alpha -\beta }\left(2\sqrt{\alpha \beta I}\right),$

where $\alpha$ and $\beta$ are determined by:

$\alpha =exp\left({\displaystyle \frac{0.49{\sigma }_R^2}{{\left(1+1.11{\sigma }_R^{\frac{12}{5}}\right)}^{\frac{5}{6}}}}\right)-1,\beta =exp\left({\displaystyle \frac{0.51{\sigma }_R^2}{{\left(1+0.69{\sigma }_R^{\frac{12}{5}}\right)}^{\frac{5}{6}}}}\right)-1.$

Here, $\alpha$ and $\beta$ signify the effective numbers of small- and large-scale turbulence cells, respectively. In these expressions, $\mathrm{\Gamma }(·)$ is the Gamma function, and $K_{\alpha -\beta }(·)$ denotes a modified Bessel function of the second kind.

The variance of intensity fluctuations, ${\sigma }_R^2$, quantifies the level of atmospheric turbulence and is given by:

(2)${\sigma }_R^2=1.23C_n^2{k}^{{\displaystyle \frac{7}{6}}}{z}^{{\displaystyle \frac{11}{6}}}$

where $C_n^2$ is the refractive index structure parameter that typically varies between ${10}^{-13}$${\mathrm{m}}^{{\displaystyle \frac{-2}{3}}}$ (strong turbulence) and ${10}^{-17}{\mathrm{m}}^{{\displaystyle \frac{-2}{3}}}$ (weak turbulence), $k={\displaystyle \frac{2\pi }{\lambda }}$ is the optical wavenumber, and z is the free-space optical (FSO) link distance. A quasi-static—or frozen—channel assumption is applied to analyze how the channel evolves over time, assuming the fading remains stable within each frame of symbols (coherence time) and updates from one frame to the next.

The impact of turbulence across different regimes, from weak to strong, can be characterized using the Rytov variance. Scintillation causes fluctuations in received signal intensity, affecting system performance. The integration of LDPC decoding plays a crucial role in mitigating these effects, significantly improving signal integrity and data reliability.

The optical power loss in the FSO channel is influenced by multiple factors, including geometric losses, atmospheric absorption, and scattering effects. The received power, $P_r$, at the destination can be expressed as:

(3)$P_r=P_t\times e^{-\alpha L}\times h_t$

where $P_t$ is the transmitted optical power, $\alpha$ is the atmospheric attenuation coefficient, L is the FSO link distance, and $h_t$ is the turbulence-induced fading factor modeled using the Gamma–Gamma distribution.

The FSO link is affected by atmospheric turbulence, which causes fluctuations in received power due to refractive index variations. The Gamma–Gamma model is widely used to characterize the probability density function (PDF) of intensity fluctuations in strong turbulence conditions. However, beyond turbulence, other impairments, such as pointing errors, beam divergence, and aperture averaging, also play a crucial role in system performance.

Pointing errors arise due to mechanical misalignment or atmospheric disturbances, leading to beam displacement at the receiver and significant power loss. The severity of pointing errors increases over long distances, contributing to additional outage probability. To mitigate this, larger receiver apertures are beneficial as they help collect more of the incoming beam and reduce signal loss. This effect, known as aperture averaging, improves system reliability by minimizing intensity fluctuations at the receiver.

Additionally, beam divergence impacts the received power, especially in longer-range FSO links. A higher divergence angle results in increased spreading loss, reducing the power density at the receiver and leading to higher BER. By optimizing the beam width and receiver aperture size, the system can effectively counteract these losses, enhancing overall link performance.

We note that our model assumes ideal alignment (no pointing error) between transmitter and receiver, which maximizes the collected optical power. In practice, pointing errors (jitter) can cause misalignment losses that compound turbulence effects [47]. To mitigate scintillation and beam wander, aperture averaging is employed in the receiver design: using a larger receiver lens helps to average out intensity fluctuations across its surface [47]. These considerations are taken into account in the link budget and design, even though the baseline model uses the Gamma–Gamma distribution for turbulence-induced fading.

Adaptive power control and spatial diversity techniques counteract transmission losses, fortifying the stability of the link. Additionally, LDPC decoding plays a crucial role in lowering the bit error rate by correcting errors that arise from atmospheric turbulence. Through its iterative approach incorporating soft decision feedback, the decoder substantially boosts error resilience, enabling the receiver to reconstruct the transmitted data accurately.

4. Results and Discussion

To assess the proposed architecture, we employed Optiwave in tandem with MATLAB R2024a-based implementations. A concise overview of the principal simulation parameters is provided in Table 1. The system operates at a data rate of 20 Gbps and supports free-space optical (FSO) links up to a distance of $2.5$ km.

We adopt an LDPC code rate of 0.5 in our simulations, which provides a strong error-correction capability suitable for turbulent FSO channels. The LDPC encoder/decoder pair uses soft-decision iterative decoding (belief propagation), with a maximum of 50 iterations allowed for convergence. The atmospheric turbulence is emulated using the Gamma–Gamma fading model described in Section 3, with a refractive index structure parameter $C_n^2$ chosen to represent strong turbulence (on the order of ${10}^{-13}$ ${\mathrm{m}}^{-2/3}$). Unless otherwise stated, we assume a link distance L = 2 km. We also include a moderate pointing loss of 1 dB in the power budget to reflect slight alignment imperfections, and the receiver aperture diameter is set to 20 cm to capture most of the beam power (aperture averaging effect). The receiver field of view is wide enough to accommodate beam divergence over 2 km without significant clipping. Taking into account optical channel degradation and receiver noise, the parameters in Table 1 ensure a realistic evaluation of the overall performance of the system.

Figure 2 illustrates the significant improvement in Bit Error Rate (BER) performance with Optical Signal-to-Noise Ratio (OSNR) when LDPC coding is applied. The dashed lines represent the original system without LDPC, where BER decreases gradually as OSNR increases but remains relatively high, especially at lower OSNR values. In contrast, the solid lines indicate the LDPC-enhanced system, which exhibits a significant reduction in BER across all measured frequencies (193.1–194.65 THz).

The improvement is evident as turbulence-induced fading leads to deep signal attenuation. Without LDPC, these fades cause severe error bursts, drastically increasing BER. However, LDPC’s iterative soft-decision decoding effectively reconstructs the original signal even in the presence of high bit error clusters, reducing BER by up to four orders of magnitude compared to uncoded transmission.

At OSNR = 12 dB, the BER improvement is relatively modest; however, as OSNR increases, the LDPC-coded system demonstrates a steeper decline in BER, reaching values as low as ${10}^{-8}$ at 32 dB OSNR, compared to ${10}^{-4}$ in the original system, demonstrating its effectiveness in combating turbulence-induced signal degradation.

The BER vs. OSNR results show a sizable coding gain at every frequency, with up to two orders of magnitude in BER improvement at higher OSNR levels. For instance, at $193.1$ THz and an OSNR of 12 dB, the uncoded BER is around ${10}^{-0.46}$, whereas the LDPC-coded BER improves to ${10}^{-0.76}$. As the OSNR rises to 32 dB, the uncoded BER at $193.1$ THz reaches ${10}^{-3.6}$, while the LDPC-coded BER pushes further to ${10}^{-5.4}$. A similar trend is observed at $194.65$ THz: at 12 dB OSNR, the uncoded BER is about ${10}^{-0.45}$, which improves to ${10}^{-0.76}$ with LDPC, and at 32 dB OSNR, it moves from ${10}^{-4.39}$ without coding to ${10}^{-7.03}$ when LDPC is applied. These results underscore LDPC’s effectiveness in markedly reducing error rates under different OSNR conditions.

Figure 3 illustrates BER performance over varying free-space optical (FSO) link distances for both the original and LDPC-coded systems. In the absence of LDPC coding (dashed lines), the BER increases as the FSO range extends from 300 m to 2200 m, highlighting the effects of optical losses and atmospheric turbulence. At 300 m, the BER for the original system at 193.1 THz is approximately ${10}^{-4}$, increasing to nearly ${10}^{-1}$ at 2200 m. A similar trend is observed across other frequencies, where, for example, at 194.65 THz, the BER degrades from approximately ${10}^{-4.5}$ at 300 m to nearly ${10}^{-1}$ at the maximum range.

Employing LDPC coding (solid lines) substantially boosts performance across all frequency channels. At a distance of 300 m, the LDPC-coded BER at $193.1$ THz reaches approximately ${10}^{-6}$, and even at 2200 m it remains under ${10}^{-3}$, indicating a coding gain of nearly two orders of magnitude. The most striking improvement occurs at higher frequencies—such as $194.65$ THz—where the LDPC-coded BER stays below ${10}^{-6}$ for most of the transmission span, in stark contrast to the uncoded scenario, which degrades severely. Uncoded transmission experiences rapid BER degradation beyond 1 km due to the combined effects of beam divergence and turbulence-induced signal fading. However, LDPC coding significantly extends the operational range of the system by mitigating deep fades. Even at 2 km, where the uncoded system exhibits a BER exceeding ${10}^{-2}$, LDPC reduces BER to below ${10}^{-5}$, ensuring reliable transmission. These findings confirm that LDPC coding effectively counters performance loss, thereby extending the maximum transmission range and ensuring stable data delivery in FSO systems.

Taken together, these outcomes confirm that LDPC coding not only strengthens error correction but also extends the operational range, ensuring more reliable data transmissions in FSO systems under varying OSNR and distance scenarios. The plotted results clearly illustrate how LDPC counteracts performance deterioration, highlighting its significance in improving both the reliability and the coverage of FSO communication links.

In addition to turbulence and noise, real-world environmental conditions can further influence FSO link performance. Adverse weather such as fog and heavy rain introduces additional attenuation and scattering, which in turn raises the required power for a given BER. For instance, recent measurements show that under clear weather, an FSO link might achieve 26 km range, but in dense fog, this range can shrink to about 1 km due to extreme optical attenuation [48]. Rain can also reduce the link span (e.g., a heavy rainfall in one study limited the FSO range to 6.7 km in a high-attenuation region) [48]. Temperature fluctuations, while not causing direct absorption, drive turbulence by creating refractive index gradients; hotter ambient conditions or urban heat can increase the strength of turbulence ($C_n^2$), leading to stronger scintillation. These environmental impairments translate to higher BER and outage probability if not accounted for. We incorporate the impact of such conditions through the Gamma–Gamma turbulence model parameters and a baseline attenuation (3 dB/km) in our simulations. Beam divergence and pointing errors also play a role: as the beam travels, it spreads, and any misalignment can result in power loss. A narrower beam (small divergence) can reduce geometric loss but is more sensitive to pointing jitter, whereas a wider beam is more robust to pointing errors but suffers higher diffraction loss. To balance this, our design uses appropriately sized transmitter optics and a large receiving aperture. A larger receiver lens not only captures more of the diverging beam power but also provides aperture averaging, which smooths out intensity fluctuations [47]. This effectively mitigates scintillation and reduces BER under turbulence. In summary, harsh environmental conditions (fog, rain, temperature-induced turbulence) will increase required link margins and can degrade BER, but careful system design—such as high transmit power, robust FEC (LDPC), and large-aperture receivers—can compensate for these effects. By designing for worst-case conditions, including some margin for pointing error and beam spreading, the WDM-PON-FSO system can maintain reliable fronthaul connectivity even in challenging weather.

The Peak-to-Average Power Ratio (PAPR) is a key factor in OFDM-based optical networks, influencing nonlinear distortions and efficiency, especially in WDM-PON-FSO systems. Its relationship with OSNR offers insights into system resilience against noise and optical impairments. As depicted in Figure 4, the PAPR vs. OSNR trends show that increasing the OSNR leads to a reduction in PAPR across every examined channel. At lower OSNR values, such as 12 dB, heightened noise levels induce waveform distortions that amplify power fluctuations, resulting in higher PAPR. For example, at $193.1$ THz, PAPR reaches roughly $10.2$ dB when the OSNR is 12 dB, while at $194.3$ THz, it stands at $9.8$ dB. As the OSNR rises to 32 dB, these values decrease to $8.5$ dB and $7.9$ dB, respectively, signifying improved signal uniformity and fewer power peaks. LDPC coding contributes to an even lower PAPR by counteracting noise-induced waveform distortions and evening out the power distribution. Post-decoding, the PAPR shows an average reduction of 1.5–2 dB at all frequencies: at $193.85$ THz, for instance, PAPR drops from $9.5$ dB at 12 dB OSNR to $7.3$ dB once LDPC is applied, and at $194.65$ THz, PAPR descends from $10.1$ dB to $7.1$ dB under high OSNR. This improvement arises from LDPC’s capacity to rectify errors and diminish high-power spikes within the signal. Interestingly, PAPR remains largely similar across each measured frequency. This uniformity stems from the inherent statistical qualities of OFDM signals, which do not significantly depend on carrier frequency. Because each WDM channel utilizes the same OFDM configuration and power adjustments, their PAPR distributions are consistent across the band. Additionally, in a linear optical chain, identical dispersion and attenuation conditions apply to all WDM signals, thereby preserving uniform PAPR behavior. These results emphasize the critical role of PAPR reduction in high-capacity optical networks and the necessity of advanced LDPC coding in WDM-PON-FSO systems.

It is worth noting that the use of OFDM in our FSO system provides certain advantages despite the primarily line-of-sight channel. OFDM enables high spectral efficiency by overlapping subcarriers and allows easy integration of multiple access techniques. Even though the FSO channel does not experience multipath frequency-selective fading in the way RF channels do, the OFDM modulation helps mitigate any potential intersymbol interference (ISI) from transmitter and receiver filtering or slight atmospheric dispersion. It also simplifies equalization, as each subcarrier can be treated as flat-fading. Moreover, using an OFDM format aligns with the 5G New Radio waveform, meaning the fronthaul can directly carry OFDM-modulated radio signals without waveform translation. This compatibility is valuable in C-RAN architectures. Therefore, the rationale for employing OFDM extends beyond diversity against multipath; it ensures maximal spectral efficiency and high data rate transmission in our WDM-PON-FSO link while maintaining robustness against any form of distortion or ISI present in the optical channel.

5. Conclusions

This work introduces a refined WDM-PON-FSO framework incorporating OFDM and LDPC decoding to address fronthaul requirements in 5G and beyond C-RAN deployments. The adoption of LDPC decoding significantly strengthens resistance to channel impairments by mitigating noise and turbulence-induced errors. Through iterative decoding, LDPC codes offer additional robustness, ensuring reliable performance even in challenging atmospheric conditions. Meanwhile, the use of OFDM boosts spectral efficiency and mitigates multipath fading effects, enabling high data rates and effective bandwidth allocation in dynamic FSO channels.

Simulation and theoretical investigations underscore LDPC’s pivotal contribution to sustaining low BER, preserving signal quality, and integrating seamlessly with WDM-PON and FSO setups. Notably, the LDPC-coded system extends the maximum reliable range to roughly 2000 m, whereas the original system experiences sharp BER deterioration past 1250–1625 m. This improvement underlines LDPC’s effectiveness in enhancing link stability over longer distances. Combined with OFDM’s adaptability and spectral efficiency, the proposed design presents a robust and scalable approach to high-capacity fronthaul networks. These findings highlight the potential of the proposed system to meet the demanding performance targets of 5G and forthcoming network generations.

Author Contributions

Conceptualization, D.A. and F.E.-N.; Methodology, D.A. and F.E.-N.; Software, D.A. and F.E.-N.; Validation, D.A. and F.E.-N.; Formal analysis, D.A. and F.E.-N.; Investigation, D.A. and F.E.-N.; Resources, D.A. and F.E.-N.; Data curation, F.E.-N.; Writing—original draft, D.A. and F.E.-N.; Writing—review & editing, F.E.-N.; Visualization, D.A. and F.E.-N.; Supervision, F.E.-N.; Project administration, F.E.-N.; Funding acquisition, D.A. and F.E.-N. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Table

Figure 1 The WDM FSO architecture.

Figure 2 BER versus OSNR plots.

Figure 3 BER versus Range plots.

Figure 4 PAPR versus OSNR plots.

Table 1

Simulation parameters.

Parameter	Value
Laser power	10 dBm
FSO attenuation	3 dB ${\mathrm{km}}^{-1}$
Beam divergence	2 mrad
Transmitter aperture diameter	5 cm
Receiver aperture diameter	20 cm
Index refraction structure ($C_n^2$)	$5\times {10}^{-17}{\mathrm{m}}^{{\displaystyle \frac{-2}{3}}}$
Spacing between WDM channels	50 GHz
Responsivity of PIN	1 A ${\mathrm{W}}^{-1}$
PIN thermal power density	$15\times {10}^{-24}$ W ${\mathrm{Hz}}^{-1}$
PIN dark current	10 nA
Gain of electrical amplifier	10 dB
Electrical amplifier power spectral density	−60 dBm ${\mathrm{Hz}}^{-1}$

References

1. Doorgakant, B.; Fowdur, T.P.; Akinsolu, M.O. End-to-End Power Models for 5G Radio Access Network Architectures with a Perspective on 6G. Mathematics; 2025; 13, 466. [DOI: https://dx.doi.org/10.3390/math13030466]

2. Ismail, S.F.; Kadhim, D.J. Towards 6G Technology: Insights into Resource Management for Cloud RAN Deployment. IoT; 2024; 5, pp. 409-448. [DOI: https://dx.doi.org/10.3390/iot5020020]

3. Wong, E. Next-Generation Broadband Access Networks and Technologies. J. Light. Technol.; 2012; 30, pp. 597-608. [DOI: https://dx.doi.org/10.1109/JLT.2011.2177960]

4. Lee, W.; Park, M.Y.; Cho, S.H.; Lee, J.; Kim, C.; Jeong, G.; Kim, B.W. Bidirectional WDM-PON based on gain-saturated reflective semiconductor optical amplifiers. IEEE Photonics Technol. Lett.; 2005; 17, pp. 2460-2462. [DOI: https://dx.doi.org/10.1109/LPT.2005.858148]

5. El-Nahal, F.; Hanik, N. Technologies for future wavelength division multiplexing passive optical networks. IET Optoelectron.; 2020; 14, pp. 53-57. [DOI: https://dx.doi.org/10.1049/iet-opt.2018.5056]

6. Khalighi, M.A.; Uysal, M. Survey on Free Space Optical Communication: A Communication Theory Perspective. IEEE Commun. Surv. Tutorials; 2014; 16, pp. 2231-2258. [DOI: https://dx.doi.org/10.1109/COMST.2014.2329501]

7. Alatawi, A.S.; Youssef, A.A.; Abaza, M.; Uddin, M.A.; Mansour, A. Effects of Atmospheric Turbulence on Optical Wireless Communication in NEOM Smart City. Photonics; 2022; 9, 262. [DOI: https://dx.doi.org/10.3390/photonics9040262]

8. Hayal, M.R.; Yousif, B.B.; Azim, M.A. Performance Enhancement of DWDM-FSO Optical Fiber Communication Systems Based on Hybrid Modulation Techniques under Atmospheric Turbulence Channel. Photonics; 2021; 8, 464. [DOI: https://dx.doi.org/10.3390/photonics8110464]

9. El-Nahal, F.; Xu, T.; AlQahtani, D.; Leeson, M. A Bidirectional Wavelength Division Multiplexed (WDM) Free Space Optical Communication (FSO) System for Deployment in Data Center Networks (DCNs). Sensors; 2022; 22, 9703. [DOI: https://dx.doi.org/10.3390/s22249703]

10. Leven, A.; Schmalen, L. Status and Recent Advances on Forward Error Correction Technologies for Lightwave Systems. J. Light. Technol.; 2014; 32, pp. 2735-2750. [DOI: https://dx.doi.org/10.1109/JLT.2014.2319896]

11. Khan, M.N.; Gilani, S.O.; Jamil, M.; Rafay, A.; Awais, Q.; Khawaja, B.A.; Uzair, M.; Malik, A.W. Maximizing Throughput of Hybrid FSO-RF Communication System: An Algorithm. IEEE Access; 2018; 6, pp. 30039-30048. [DOI: https://dx.doi.org/10.1109/ACCESS.2018.2840535]

12. Anandkumar, D.; Sangeetha, R.G. Performance Evaluation of LDPC-Coded Power Series Based Malaga (M) Distributed MIMO/FSO Link With M-QAM and Pointing Error. IEEE Access; 2022; 10, pp. 62037-62055. [DOI: https://dx.doi.org/10.1109/ACCESS.2022.3180835]

13. Galijasevic, S.; Luo, J.; Divsalar, D.; Wesel, R. Channel-Prediction-Driven Rate Control for LDPC Coding in a Fading FSO Channel with Delayed Feedback. IEEE Open J. Commun. Soc.; 2025; early access [DOI: https://dx.doi.org/10.1109/OJCOMS.2025.3538839]

14. Youssef, A.A.; Elsanadily, S.I. Enhancing free-space optical communication networks using Generalized Low Density Parity Check codes. Opt. Laser Technol.; 2025; 181, 111862. [DOI: https://dx.doi.org/10.1016/j.optlastec.2024.111862]

15. Gueye, I.; Diop, I.; Dioum, I.; Keita, K.W.; Ndiaye, P.; Diallo, M.; Farssi, S.M. Performance Analysis of Mixed MIMO RF/FSO DF Relaying Based on Globally Coupled Low Density Parity Check (GC-LDPC) Codes. Proceedings of the 2021 23rd International Conference on Advanced Communication Technology (ICACT); PyeongChang, Republic of Korea, 7–10 February 2021; pp. 14-22. [DOI: https://dx.doi.org/10.23919/ICACT51234.2021.9370871]

16. Shi, Y.; Armghan, A.; Ali, F.; Aliqab, K.; Alsharari, M. Enriching Capacity and Transmission of Hybrid WDM-FSO Link for 5G Mobility. Photonics; 2023; 10, 121. [DOI: https://dx.doi.org/10.3390/photonics10020121]

17. El-Nahal, F.; Xu, T.; AlQahtani, D.; Leeson, M. A Bidirectional WDM-PON Free Space Optical (FSO) System for Fronthaul 5 G C-RAN Networks. IEEE Photonics J.; 2023; 15, 7300208. [DOI: https://dx.doi.org/10.1109/JPHOT.2022.3232081]

18. Yeh, C.H.; Wang, B.Y.; Hsu, W.H.; Liu, L.H.; Ko, H.S. A Simple WDM-PON Architecture Together With Private Interconnected ONUs. IEEE Access; 2021; 9, pp. 126319-126323. [DOI: https://dx.doi.org/10.1109/ACCESS.2021.3110729]

19. AlQahtani, D.; El-Nahal, F. Coherent WDM-PON and free space optical (FSO) system for front-haul in next-generation cellular networks. Optik; 2025; 323, 172212. [DOI: https://dx.doi.org/10.1016/j.ijleo.2024.172212]

20. Park, S.J.; Lee, C.H.; Jeong, K.T.; Park, H.J.; Ahn, J.G.; Song, K.H. Fiber-to-the-home services based on wavelength-division-multiplexing passive optical network. J. Light. Technol.; 2004; 22, pp. 2582-2591. [DOI: https://dx.doi.org/10.1109/JLT.2004.834504]

21. Torres-Ferrera, P.; Effenberger, F.; Faruk, M.S.; Savory, S.J.; Gaudino, R. Overview of high-speed TDM-PON beyond 50 Gbps per wavelength using digital signal processing [Invited Tutorial]. J. Opt. Commun. Netw.; 2022; 14, pp. 982-996. [DOI: https://dx.doi.org/10.1364/JOCN.468920]

22. Zhang, J.; Jia, Z. Coherent Passive Optical Networks for 100G/λ-and-Beyond Fiber Access: Recent Progress and Outlook. IEEE Netw.; 2022; 36, pp. 116-123. [DOI: https://dx.doi.org/10.1109/MNET.005.2100604]

23. Feng, N.; Ma, M.; Zhang, Y.; Tan, X.; Li, Z.; Li, S. Key Technologies for a Beyond-100G Next-Generation Passive Optical Network. Photonics; 2023; 10, 1128. [DOI: https://dx.doi.org/10.3390/photonics10101128]

24. Ullah, R.; Liu, B.; Zhang, Q.; Tian, Q.; Tian, F.; Qu, Z.; Yan, C.; Khan, M.S.; Ahmad, I.; Xin, X. Optical frequency comb generation with high tone-to-noise ratio for large-capacity wavelength division multiplexed passive optical network. Opt. Eng.; 2015; 54, 116102. [DOI: https://dx.doi.org/10.1117/1.OE.54.11.116102]

25. Effenberger, F.J.; Zhang, D. WDM-PON for 5G Wireless Fronthaul. IEEE Wirel. Commun.; 2022; 29, pp. 94-99. [DOI: https://dx.doi.org/10.1109/MWC.001.2100420]

26. Vujicic, Z.; Santos, M.C.; Méndez, R.; Klaiqi, B.; Rodriguez, J.; Gelabert, X.; Rahman, M.A.; Gaudino, R. Toward Virtualized Optical-Wireless Heterogeneous Networks. IEEE Access; 2024; 12, pp. 87776-87806. [DOI: https://dx.doi.org/10.1109/ACCESS.2024.3417358]

27. Youssef, A.A.; Abaza, M.; Alatawi, A.S. LDPC Decoding Techniques for Free-Space Optical Communications. IEEE Access; 2021; 9, pp. 133510-133519. [DOI: https://dx.doi.org/10.1109/ACCESS.2021.3116081]

28. Gueye, I.; Diop, I.; Dioum, I. Performance analysis of multi-user mixed RF and hybrid RF/FSO cooperative systems with buffers based on GC-LDPC codes. Proceedings of the 2021 IEEE 2nd International Conference on Signal, Control and Communication (SCC); Tunis, Tunisia, 20–22 December 2021; pp. 155-160. [DOI: https://dx.doi.org/10.1109/SCC53769.2021.9768382]

29. Kose, C.; Halford, T.R. Incremental redundancy hybrid ARQ protocol design for FSO links. Proceedings of the MILCOM 2009–2009 IEEE Military Communications Conference; Boston, MA, USA, 18–21 October 2009; pp. 1-7. [DOI: https://dx.doi.org/10.1109/MILCOM.2009.5380093]

30. Gao, S.; Liu, X.; Chen, Y.; Jiang, J.; Liu, Y.; Jiang, Y. Atmospheric Turbulence Strength Estimation Using Convolution Neural Network. IEEE Photonics J.; 2023; 15, pp. 1-7. [DOI: https://dx.doi.org/10.1109/JPHOT.2023.3314833]

31. Jia, C.; Wang, P.; Li, Y.; Huang, C.; Fu, H.; Pang, W. ABERs of LDPC-Coded Multi-Hop FSO Over Double GG Fading Channels With Pointing Error and Path Loss. IEEE Photonics Technol. Lett.; 2018; 30, pp. 1357-1360. [DOI: https://dx.doi.org/10.1109/LPT.2018.2846615]

32. Barua, B.; Majumder, S.P. Analytical Performance Evaluation of an LDPC Coded FSO Communication System under Strong Atmospheric Turbulence. Proceedings of the 2018 International Conference on Smart Systems and Inventive Technology (ICSSIT); Tirunelveli, India, 13–14 December 2018; pp. 607-611. [DOI: https://dx.doi.org/10.1109/ICSSIT.2018.8748402]

33. Djordjevic, I.B.; Djordjevic, G.T. On the high-speed communication over hybrid free-space optical (FSO)-wireless fading channels. Proceedings of the 2009 IEEE LEOS Annual Meeting Conference Proceedings; Belek-Antalya, Turkey, 4–8 October 2009; pp. 833-834. [DOI: https://dx.doi.org/10.1109/LEOS.2009.5343426]

34. Youssef, A.A. Bootstrapped low complex iterative LDPC decoding algorithms for free-space optical communication links. EURASIP J. Wirel. Commun. Netw.; 2023; 2023, 78. [DOI: https://dx.doi.org/10.1186/s13638-023-02285-w]

35. Hsu, C.H.; Jiang, S.Y.; Hsieh, S.E.; Yeh, C.H.; Lai, Y.T.; Chen, L.Y.; Liaw, S.K.; Chow, C.W. Hybrid Self-Protected Fiber-FSO WDM-PON System with Fiber Breakage Prevention. Photonics; 2022; 9, 822. [DOI: https://dx.doi.org/10.3390/photonics9110822]

36. Yeh, C.H.; Ko, H.S.; Liaw, S.K.; Liu, L.H.; Chen, J.H.; Chow, C.W. A Survivable and Flexible WDM Access Network by Alternate FSO- and Fiber-Paths for Fault Protection. IEEE Photonics J.; 2022; 14, 7209205. [DOI: https://dx.doi.org/10.1109/JPHOT.2021.3140095]

37. Yeh, C.H.; Lin, W.P.; Jiang, S.Y.; Hsieh, S.E.; Hsu, C.H.; Chow, C.W. Integrated Fiber-FSO WDM Access System with Fiber Fault Protection. Electronics; 2022; 11, 2101. [DOI: https://dx.doi.org/10.3390/electronics11132101]

38. Alhalabi, M.; Taşpınar, N.; El-Nahal, F.I. Bidirectional intensity modulated/direct detection optical OFDM WDM-PON system. Optoelectronics Lett.; 2023; 19, pp. 83-87. [DOI: https://dx.doi.org/10.1007/s11801-023-2126-x]

39. Wang, Z.; Xu, Z.; Chen, K.; Qu, Y.; Liu, X.; Li, Y.; Zhang, J. Efficient Pipeline Conflict Resolution for Layered QC-LDPC Decoders in OFDM-PON. Photonics; 2024; 11, 429. [DOI: https://dx.doi.org/10.3390/photonics11050429]

40. Cowley, W.G.; Lechner, G.; Biggs, C.; Grant, K.J. Transceiver design for reliable free space optical communications. Proceedings of the 2011 Australian Communications Theory Workshop; Melbourne, VIC, Australia, 31 January 2011–3 February 2011; pp. 94-99. [DOI: https://dx.doi.org/10.1109/AUSCTW.2011.5728744]

41. Zhao, X.; Chen, Y.; He, H.; Zhou, L.; Dong, Z. LDPC-Coded Probabilistic Shaping PAM8 Employing a Novel Bit-Weighted Distribution Matching in WDM-PON. J. Light. Technol.; 2020; 38, pp. 4641-4647. [DOI: https://dx.doi.org/10.1109/JLT.2020.2994387]

42. He, H.; Chen, Y.; Xiao, Q.; Zhou, L.; Dong, Z. LDPC-Coded Probabilistic Shaping PAM4 Based on Many-to-One Mapping in WDM-PON. J. Light. Technol.; 2020; 38, pp. 3918-3925. [DOI: https://dx.doi.org/10.1109/JLT.2020.2984828]

43. Xiao, Q.; Chen, Y.; Lin, S.; He, H.; Wu, X.; You, J.; Zeng, Y.; Zhou, L.; Dong, Z. DFT-Spread DMT-WDM-PON Employing LDPC-Coded Probabilistic Shaping 16 QAM. J. Light. Technol.; 2020; 38, pp. 714-722. [DOI: https://dx.doi.org/10.1109/JLT.2019.2948145]

44. Liu, X.; Wang, P.; Liu, T.; Li, Y.; Guo, L.; Tian, H. ABER Performance of LDPC-Coded OFDM Free-Space Optical Communication System Over Exponentiated Weibull Fading Channels With Pointing Errors. IEEE Photonics J.; 2017; 9, 7905113. [DOI: https://dx.doi.org/10.1109/JPHOT.2017.2723427]

45. Al-Habash, M.; Andrews, L.C.; Phillips, R.L. Mathematical model for the irradiance probability density function of a laser beam propagating through turbulent media. Opt. Eng.; 2001; 40, pp. 1554-1562. [DOI: https://dx.doi.org/10.1117/1.1386641]

46. Wang, Z.; Zhong, W.D.; Fu, S.; Lin, C. Performance comparison of different modulation formats over free-space optical (FSO) turbulence links with space diversity reception technique. IEEE Photonics J.; 2009; 1, pp. 277-285. [DOI: https://dx.doi.org/10.1109/JPHOT.2009.2039015]

47. Elsayed, E.E.; Hayal, M.R.; Nurhidayat, I.; Shah, M.A.; Elfikky, A.; Boghdady, A.I.; Juraev, D.A.; Morsy, M.A. Coding techniques for diversity enhancement of dense wavelength division multiplexing MIMO-FSO fault protection protocols systems over atmospheric turbulence channels. IET Optoelectron.; 2024; 18, pp. 11-31. [DOI: https://dx.doi.org/10.1049/ote2.12111]

48. Singh, M.; Elsayed, E.E.; Alayedi, M.; Aly, M.H.; Abd El-Mottaleb, S.A. Performance analysis in spectral-amplitude-coding-optical-code-division-multiple-access using identity column shift matrix code in free space optical transmission systems. Opt. Quantum Electron.; 2024; 56, 795. [DOI: https://dx.doi.org/10.1007/s11082-023-05721-9]