Free-space optical communication (FSO) can provide high-capacity, rapidly deployable links without a physical fiber connection. It is used in applications such as last-mile access, network-node interconnection, fiber backup and wireless backhaul. But unlike a protected fiber path, an FSO link passes through the atmosphere, where turbulence can produce rapid fluctuations in received optical power. These fluctuations, known as scintillation or fading, can reduce link reliability and interrupt transmission when the received signal falls below the receiver’s sensitivity.
For engineers and system designers, managing turbulence is not a single-component problem. Options include optical amplification, spatial or temporal diversity, larger receiver apertures, and error-correcting codes. Their suitability depends on the channel conditions, link architecture and receiver design.
How turbulence affects an FSO link
Atmospheric turbulence creates time-varying changes in the air’s refractive index. As an optical beam propagates, those changes can cause fluctuations in irradiance at the receiver. The impact is particularly relevant over longer paths, including distances around a kilometre or more, and can affect both link capacity and latency. In severe fades, the received power may drop below the level needed for reliable detection.
Channel models help describe the expected fluctuations and evaluate mitigation strategies. The scintillation index, a normalized measure of irradiance variance, is one indicator of fading strength. Log-normal distributions are commonly used to represent weak turbulence, while Gamma-Gamma and other distributions are used to model moderate-to-strong conditions. These are statistical descriptions, not guarantees of performance: actual link assessment needs to reflect the path and system assumptions being considered.
Free-space optical communication turbulence mitigation options
Optical amplification and signal equalization
A saturated semiconductor optical amplifier (SOA) or an erbium-doped fiber amplifier (EDFA) can amplify the received optical signal and reduce intensity fluctuations. The reported performance depends on turbulence severity and receiver configuration. Under lower turbulence in one laboratory study, a saturated amplifier or EDFA suppressed much of the scintillation; at higher turbulence levels, fluctuations were not eliminated. The EDFA showed better bit-error-rate performance than the tested SOA configurations in those conditions.
Amplification can therefore improve receiver sensitivity and help limit the effect of fading, but it should not be treated as a complete remedy. System designers need to assess the amplifier as part of the full optical and detection chain, including the expected channel conditions and required error performance.
Spatial diversity and combining
Spatial diversity uses multiple transmit apertures, receive apertures, or both. Because separated paths can experience different fading, combining their signals can reduce the likelihood that a single deep fade disrupts the whole link. Apertures are typically separated enough to help make their received fading less correlated, although correlation can limit the diversity benefit.
At the receiver, common approaches include:
- Selection combining: uses the branch with the strongest received signal or signal-to-noise ratio.
- Equal-gain combining: combines branches with equal weighting, offering a less complex alternative to optimal weighting.
- Maximal-ratio combining: weights branches to maximize the output signal-to-noise ratio.
The choice involves a trade-off between implementation complexity and combining performance. Diversity can also be paired with modulation approaches such as subcarrier intensity modulation. However, system-level benefits depend on channel independence and the selected detection and combining methods.
Aperture averaging
Aperture averaging reduces the effect of small-scale intensity variation by collecting light across a receiver aperture rather than at a point. The aperture-averaging factor compares the normalized irradiance variance for a finite receiver diameter with the variance for a point receiver. Increasing the aperture relative to the atmospheric coherence scale can reduce measured scintillation, though the actual effect depends on propagation distance and other optical parameters.
Coding and time diversity
Channel coding adds redundancy that enables a receiver to detect or correct errors. Interleaving can spread transmitted symbols across time, helping the system benefit from time diversity when fading changes over the transmission interval. Reviewed approaches include convolutional, Reed-Solomon, concatenated and turbo codes, as well as low-density parity-check (LDPC) and rateless codes.
Different schemes suit different conditions and decoding requirements. The reviewed work identifies convolutional coding as a potential fit for weak turbulence when decoding complexity matters, while turbo coding may be more appropriate in strong turbulence. Interleaving must also be sized with channel correlation in mind: if fades persist for long periods relative to symbol duration, very large interleavers may be impractical. OFDM and rateless coding are among approaches considered for addressing demanding channel conditions.
Architecture choices: relays and channel assumptions
Relays can extend FSO coverage through serial, multi-hop links or parallel cooperative configurations. In decode-and-forward (DF), a relay detects and decodes the incoming signal before retransmitting it. In amplify-and-forward (AF), it amplifies and forwards the received signal. For a serial DF route, an outage on any intermediate link can cause an end-to-end outage, so each hop must be considered in the link budget and reliability analysis.
Model assumptions matter, too. Far-field approximations can simplify analysis, but may not accurately represent every practical geometry. Diffraction and spatial correlation can reduce diversity gains or introduce crosstalk in multiplexed systems. Near-field modeling can better represent configurations such as compact multi-beam transmitters or large receive apertures, although it may require numerical methods rather than a closed-form solution.
Practical implications for FSO system design
Mitigation strategies address different parts of the problem: amplification acts on received signal strength and fluctuations, diversity provides alternate optical paths, aperture averaging reduces spatial intensity variation, and coding adds resilience to detection errors. Relays address coverage, while channel models help estimate performance under specified turbulence assumptions.
Engineers should evaluate these techniques against the intended path, receiver architecture, channel correlation, complexity and target error performance. No single method removes every effect of turbulence. A robust design starts with a suitable channel model and combines mitigation measures where their benefits match the system’s requirements. Optical safety also remains part of the design, particularly when selecting laser sources and operating wavelengths.