The work, carried out by Abdullah Oran, builds a physics-informed numerical model of a complete coherent optical receiver, connecting three previously separate threads into one framework: the near-field collection gain that experiments had measured but never fully explained, the loss that the lantern itself introduces, and the statistical diversity that its multiple output branches provide. The result is not just a simulation that reproduces known data. It is a quantitative dissection of the mechanisms at play, complete with falsifiable predictions that can be tested on a laboratory bench.

The starting point is the well-known fragility of single-mode fiber coupling. Coupling efficiency is computed as the overlap between the received optical field and the fiber’s fundamental Gaussian mode, and this overlap is brutally sensitive to mismatch. When a transmitter sits close to the receiver—inside the so-called Fraunhofer distance, beyond which a beam has developed its far-field shape—the light arriving at the fiber plane is larger and more curved than the fiber mode expects. For the receiver modeled here, with a 2.1-millimeter effective aperture at 1550 nanometers, that critical distance is about 5.69 meters. Inside it, the spot delivered to the fiber swells dramatically: at half a meter, the model predicts an effective spot of roughly 59 micrometers on the single-mode path, more than ten times the 5.2-micrometer mode-field radius of the fiber. The overlap integral collapses, and with it the received power.

This is where the photonic lantern enters. The device is a continuous taper that transforms a multimode input—in this case a 50-micrometer core with a numerical aperture of 0.22—adiabatically into nineteen single-mode outputs. Light that would overfill and miss the single-mode mode field entirely is instead caught by the multimode input and distributed across the branches. The model captures this asymmetry through two spot-growth exponents, fixed in advance rather than fitted: the single-mode path’s spot grows with distance as a full power law, while the multimode path, far more tolerant, grows more gently. At half a meter the lantern-side spot is about 18 micrometers, still comfortably inside the 25-micrometer core radius, so the lantern keeps collecting power while the single-mode fiber loses it.

The model’s most striking finding comes from switching individual mechanisms off. When the differential spot growth is disabled, the predicted near-field gain of roughly 8 decibels not only vanishes but reverses sign, dropping to about minus 1.5 decibels. The gain, in other words, is not an amplification property of the lantern at all—it is a symptom of the single-mode fiber’s vulnerability. Removing the residual wavefront curvature term costs only a few tenths of a decibel, marking it as a genuine but secondary contributor. Meanwhile, disabling the lantern’s own mode-overflow penalty raises the predicted gain to nearly 16 decibels, revealing that the device’s finite port count actually imposes a several-decibel cost in the near field: when the input excites more spatial modes than the lantern has ports, the excess power is radiated away in the taper. At half a meter the model estimates around 115 excited modes against only 19 ports, producing roughly 9 decibels of transition loss.

Perhaps the most consequential result concerns what happens beyond the near field. As propagation distance increases, the average collection gain decays monotonically toward zero—by ten meters the two receivers collect nearly identical mean power. A naive reading would suggest the lantern’s advantage has evaporated. The diversity statistics say otherwise. Because the nineteen branches carry partially decorrelated speckle fluctuations, coherently combining them with maximum-ratio combining compresses the fluctuations in the received signal-to-noise ratio by a factor of about 5.4 in the logarithmic domain, even at ten meters where the mean advantage is a negligible 0.25 decibels. The probability of a deep fade below a minus 5-decibel threshold drops from 0.37 for direct single-mode detection to below one in a thousand. Outage probability, the metric that ultimately determines whether a link is usable, improves by orders of magnitude.

The correlation structure underlying this diversity benefit is itself distance-dependent. In the near field, where the received field has limited spatial complexity, the lantern branches are moderately correlated, with a coefficient of about 0.55. As the beam propagates and the speckle pattern decorrelates, the coefficient falls toward a residual of 0.05 with a decorrelation distance of roughly one meter. Less correlated branches mean more independent looks at the fading channel, which is exactly why the diversity advantage persists—and even strengthens—at distances where the average power gain has disappeared. This is consistent with earlier experimental observations from lantern-based coherent LIDAR and photon-counting receiver studies, though the author is careful to note that these comparisons remain qualitative.

Unusually for a modeling paper, every numerical component is verified against closed-form analytical results. The speckle synthesizer reproduces the unit intensity contrast of fully developed speckle to within 0.4 percent; the overlap-integral engine matches analytical Gaussian coupling efficiencies to five significant digits; the maximum-ratio combiner reproduces the analytical variance and Erlang-distributed outage of independent branches to within about 2 percent. The author also quantifies what the calibration does not guarantee: because only two experimental anchor points exist—an 8-decibel gain at 0.5 meters and near-unity gain at 10 meters—the fitted parameters are not unique, and all near-optimal parameter sets are retained as an uncertainty band spanning roughly half a decibel. Sensitivity analysis shows the entire near-field prediction hinges on a single exponent pair fixed a priori, which is precisely where the model is most exposed to experimental test.

That test is spelled out in detail. The study proposes a bench-scale validation protocol built around five measurements: imaging the effective spot size versus distance with a beam profiler, measuring wavefront defocus with a Shack–Hartmann sensor, recording the power distribution across the lantern’s nineteen ports, detecting pairs of branches coherently to extract inter-branch correlation, and repeating the full gain-versus-distance curve with fading statistics. Each measurement comes with a quantitative prediction and an explicit falsification criterion—for example, if the measured single-mode spot slope near 0.5 meters is close to minus 0.5 rather than minus 1.0, the dominant mechanism proposed here would be refuted. The two highest-value measurements require nothing more exotic than power meters, a beam profiler and the lantern itself.

The practical significance extends across the growing portfolio of free-space optical applications, from satellite downlinks to drone links and short-range interconnects, where coherent detection and digital signal processing demand stable single-mode coupling. The study reframes the photonic lantern not as a magic amplifier but as a mode-mismatch insurance policy: it pays a modest, quantifiable premium in transition loss and insertion loss, and in exchange it protects the link against the near-field coupling collapse and against the deep speckle fades that plague single-mode reception at all distances. By converting a phenomenological observation into a small set of measurable, refutable statements, the model gives experimentalists a concrete roadmap—and gives system designers a principled basis for deciding when a nineteen-branch lantern is worth its 1.3 decibels of baseline insertion loss.

Subject of Research: Physics-informed modeling of photonic lantern collection gain and spatial diversity in coherent free-space optical receivers

Article Title: Physics-informed modeling of photonic lantern collection gain and diversity in free-space optical reception

Article References: Oran, A. (2026). Physics-informed modeling of photonic lantern collection gain and diversity in free-space optical reception. Results in Optics, 25, Article 101173. https://doi.org/10.1016/j.rio.2026.101173

Image Credits: AI Generated

DOI: 10.1016/j.rio.2026.101173

Keywords: photonic lantern, free-space optical communication, single-mode fiber coupling, spatial diversity, maximum-ratio combining, speckle, outage probability, wavefront curvature, mode overflow loss, coherent optical reception, Fraunhofer distance, Monte Carlo simulation

The research, led by Nikita Goel and Pankaj Kumar of Manipal Institute of Technology together with Vrinda Gupta of the National Institute of Technology Kurukshetra, tackles a scenario known as drone-assisted cooperative communication, or DACC. In such systems, a source node on the ground cannot reach the destination directly with sufficient quality, so a drone hovering between them receives the signal, strengthens it, and retransmits it. Because the drone is the linchpin of the link, draining its battery quickly collapses the entire connection, which is why the authors focused their energy-harvesting design on the relay itself.

What distinguishes this work from earlier energy-harvesting schemes is the way the drone decides how much of each received signal to divert into its battery. Conventional designs use a fixed power-splitting factor, dividing every incoming signal by a constant fraction regardless of conditions. The researchers instead introduce a statistical, height-dependent splitting factor. At any given altitude and set of environmental parameters, the drone computes the probability that a line-of-sight path exists between itself and the ground node, then sets its scavenging ratio accordingly. When the line-of-sight probability is high and received power is strong, the drone banks more energy; when the path is obstructed, it shifts the balance back toward information transmission.

The physics behind that decision rests on the air-to-ground channel model. Rather than assuming the idealized extremes used in much of the prior literature, the team modeled the link between the drone and ground users with Nakagami-m fading, a flexible statistical model that can capture mixtures of line-of-sight and non-line-of-sight propagation. The direct ground link between source and destination, assumed to be purely non-line-of-sight in the dense scenario they study, uses Rayleigh fading. Crucially, because the drone moves vertically, the channels in this hybrid environment are not independent: the correlation between the source-drone, drone-destination, and direct links varies with altitude, and the mathematical framework explicitly tracks this height-dependent correlation.

Within this correlated hybrid fading environment, the researchers derived closed expressions for two key performance metrics: outage probability, the chance that the link fails to deliver a target data rate, and achievable rate at the destination. The destination node combines the direct signal received in the first time slot with the relayed signal received in the second using maximum ratio combining, extracting the best of both paths. The drone can operate in either amplify-and-forward mode, which scales and retransmits the analog received signal, or decode-and-forward mode, which decodes, re-encodes, and retransmits it. Two algorithms were developed: one governing the scavenging and information-splitting decisions at the drone, and another computing rate and outage probability across the correlated channel environment, implemented in MATLAB simulations.

The simulation results reveal a nuanced trade-off. Outage probability rises with drone altitude in every scenario considered, an effect the authors attribute to the Nakagami-m shaping parameters being treated as height-independent, so that path loss eventually dominates any line-of-sight gain. When the drone transmits using only the power it has harvested, performance is worst, because the harvested energy fluctuates with channel conditions from one transmission session to the next. The best configuration lets the drone transmit at a fixed, relatively high power while simultaneously scavenging energy to sustain it, effectively replenishing the battery that would otherwise deplete steadily.

That sustained battery translates directly into longevity. Compared with a system in which the drone draws all relay power from its primary battery, the power-scavenging scheme completes significantly more communication cycles for the same initial charge, and substantially more packets arrive successfully at the destination. Although harvesting does slightly worsen outage performance in some regimes, because power siphoned into the battery is unavailable for retransmission, the authors show that this drawback is outweighed by the dramatic increase in total delivered data over the life of the network. In a crisis scenario, that difference is measured not in abstract metrics but in the number of messages that get through before the aerial relay falls silent.

Environment matters as well. The team compared outage behavior across dense urban, urban, and suburban settings, finding the worst results in dense urban terrain, where tall buildings suppress the probability of a line-of-sight connection and depress the received signal-to-noise ratio, while suburban environments, with clearer sightlines, performed best. The height-dependent splitting factor adapts across all of these contexts, adjusting scavenging intensity to the environment-specific line-of-sight probability in a way that a static design cannot.

The study also contributes a sharper account of channel correlation than most prior drone-relay analyses. At low altitudes, the source-drone and source-destination links, as well as the drone-destination and direct links, exhibit strong correlation because the geometry of the moving drone couples them; as the drone climbs, that coupling weakens and different link pairs take on the stronger relationship. Because most of the literature treats relay channels as fixed and independent, this height-driven correlation dynamic has been largely overlooked, even though it materially affects outage and rate predictions in real deployments.

The authors position the work within the march toward 5G and beyond-5G networks, where users demand higher throughput, better reliability, and lower energy consumption from battery-constrained devices, and where drones are increasingly folded into cooperative communication architectures. They note that the scheme applies to infrastructure-less wireless networks generally, and that extending it to multi-user scenarios and deriving closed-form performance expressions are the next research steps. For disaster response teams weighing how long an aerial relay can keep a shattered network breathing, the message is that the drone’s own下行 data stream can double as a fuel line, and that tuning how much of that stream to bottle up, altitude by altitude, can stretch mission endurance considerably.

Subject of Research: A height-dependent radio-frequency energy harvesting scheme for drone-assisted cooperative communication in correlated hybrid fading channels.

Article Title: Power Scavenging for Strengthening the Life Cycle of Cooperative Devices in Correlated Hybrid Fading Environment

Article References: Goel, N., Gupta, V., & Kumar, P. (2026). Power Scavenging for Strengthening the Life Cycle of Cooperative Devices in Correlated Hybrid Fading Environment. Mobile Networks and Applications. https://doi.org/10.1007/s11036-026-02526-4

Image Credits: AI Generated

DOI: 10.1007/s11036-026-02526-4

Keywords: drone relays, energy harvesting, cooperative communication, Nakagami-m fading, channel correlation, outage probability, power splitting, UAV communications, network lifetime, wireless networks, 5G, disaster response