Free-space optical communications can require fast, accurate beam steering between moving platforms while maintaining enough optical gain to establish a reliable link. A single large telescope can provide a substantial aperture, but its size and mass complicate rapid movement, while damage to the aperture can create a single point of failure. A phased array of phased arrays (PAPA) offers a different approach: combine multiple smaller apertures and control their optical phases so they act together as a transmitter or receiver.
The concept is especially relevant where platform constraints, atmospheric turbulence and link acquisition all influence system design. Liquid crystal on silicon (LCoS) optical phased arrays can steer beams without mechanically moving the aperture segments. The architecture also introduces trade-offs in alignment, aperture efficiency, response time and field of regard.
Why use a segmented aperture for free-space optical communications?
Large optical assemblies can be difficult to position quickly and accurately. They may also compete with other platform requirements for space and uninterrupted window area. Dividing the aperture into smaller elements can make placement more flexible and reduce the consequences of damage to any one part: the remaining segments may still contribute to the link.
Segmentation can also help address atmospheric effects. When an individual sub-aperture is smaller than the spatial coherence length of the received light, turbulence is less likely to diminish the signal across that element. If the segments are properly phased, their contributions can combine coherently to increase receiver gain in proportion to total aperture area. Without proper phasing, the gain does not add in the same way; the paper describes an increase proportional to the square root of the number of array elements.
This distinction matters in changing conditions. Under weak turbulence, phase differences across the aperture may vary slowly enough for coherent combining to be practical. In stronger or more dynamic turbulence, maintaining alignment between segments can be difficult. Noncoherent combining can avoid some of the problems associated with interference at a detector, although it does not provide the same coherent gain.
How an LCoS phased array of phased arrays works
An optical phased array (OPA) steers light by applying a spatial phase pattern across a beam’s wavefront. A linear phase ramp changes the beam’s propagation angle; a spherical phase pattern can make the device function as a programmable lens. Because phase is cyclic, a ramp can use repeated 2π resets rather than requiring an indefinitely increasing phase shift.
In the PAPA arrangement, multiple small beams are controlled by individual OPAs. The described LCoS design pairs X- and Y-steering elements to provide two-dimensional control. Arranged in thin, stacked elements resembling Venetian blinds, the paired steering stages direct each beam while allowing the overall output to remain approximately in line with the system’s optical axis. The paper calls this an in-line “transmissive” arrangement, distinguishing it from the reflection-mode operation associated with the LCoS devices themselves.
A fiber feed and collimator deliver beams to the OPA elements. For receiving or alignment, a small portion of the steered light can be sampled and directed to a detector. Focal-plane measurements provide information used to phase the array as a common aperture. Each segment can be adjusted in tip, tilt and optical path, enabling the system to combine beams in the far field.
Beam acquisition, tracking and combining
Non-mechanical steering avoids some control issues associated with moving optical assemblies, such as overshoot and ringing. The applied phase pattern sets the steering direction relative to the optical axis, so the approach does not rely on positional feedback from a mechanical steering mechanism to determine the beam angle.
The ability to change the phase pattern also supports beam shaping. A PAPA system can broaden a beam or form multiple spots to search a wider area during acquisition. As the link is established, the distortion can be reduced to concentrate energy and improve link margin. This offers a way to move from broader coverage toward a narrower beam without relying solely on conventional mechanical scanning.
For a receiver, the array can combine signals coherently when phase alignment is achievable or noncoherently when environmental conditions make it difficult. The appropriate approach depends on turbulence and the system’s ability to detect and respond to wavefront changes. Adaptive correction requires measurement and processing, which can limit how quickly the system responds to rapidly changing distortions.
Engineering constraints and design trade-offs
Fill factor and aperture efficiency
The proportion of the overall aperture occupied by active optical area affects how efficiently the array uses its physical size. In the design examples discussed in the paper, an OPA arrangement had a 59% fill factor before fiber-feed constraints were included. Accounting for the collimator and alignment space reduced the estimated array fill factor to approximately 50%. These figures describe the particular design examined, not a universal PAPA performance level.
Packaging is not determined only by the OPA dimensions. Electrical connections, louver thickness, the 45-degree operating angle and the size of the fiber-feed collimators all influence spacing. A sparse aperture can also distribute transmitted energy into sidelobes, so fill factor is relevant to both transmitter beam quality and receiver aperture efficiency.
Switching transients and field of regard
Changing a phase pattern can temporarily reduce power in the main beam and create an unwanted on-axis component. The paper describes a roughly 50% main-lobe intensity drop for small-angle switching in the devices under discussion. Possible approaches included updating sub-patterns in sequence, using multiple beam lobes to maintain tracking, or using greater phase modulation to adjust the steering slope more continuously. Each approach carries a trade-off, such as slower scanning, reduced power at the receiver or some beam-profile distortion.
The steering range is another constraint. The paper reports that the OPA discussed could steer to about ±3 degrees at 1.55 micrometers before efficiency fell below 50%. Wider coverage would require improvements to the OPA or additional optical steering approaches. The authors considered larger active-area devices and other ways of extending the field of regard, but these were development directions rather than established results in the described system.
Practical implications for system designers
A phased array of phased arrays is not simply a way to replace one large telescope with several small ones. Its value depends on coordinating the segments, managing phase alignment, fitting the fiber and electrical assemblies into the available aperture, and matching steering behavior to link dynamics. Turbulence conditions influence whether coherent combining is realistic, while switching response and field of regard affect acquisition and tracking.
For engineers evaluating free-space optical communications, the central design question is whether distributed, electronically controlled apertures address the platform’s constraints enough to justify their added integration and control demands. PAPA architecture offers potential advantages in placement flexibility, beam shaping and resilience to localized aperture damage. Realizing those benefits requires attention to fill factor, atmospheric coherence, transient behavior and steering range throughout the system design.