Inter-Satellite Links: How LEO Satellites Are Becoming a Network in Space

From bent-pipe relays to optical meshes, inter-satellite links are changing the architecture of satellite communications.

For most of the satellite era, communications spacecraft behaved largely as relays between two points on Earth. A signal travelled from a ground terminal to the satellite, was processed or amplified, and then returned through another ground station. Engineers often described this architecture as a bent-pipe system because the satellite mainly redirected traffic rather than deciding where that traffic should go.

That model remains effective in many applications, particularly in geostationary satellite systems. But large low Earth orbit constellations have introduced a very different networking problem.

A LEO satellite moves rapidly across the Earth. Its view of users, gateways, neighbouring spacecraft, and terrestrial infrastructure changes continuously. Over oceans, deserts, polar regions, and other remote areas, the satellite may have a perfectly good connection to a customer while having no suitable gateway beneath it.

Inter-satellite links change this limitation fundamentally.

Instead of forcing every packet to return immediately to Earth, one satellite can send the traffic directly to another spacecraft. That satellite can forward it again, and the packet can remain inside the constellation until the network identifies the most appropriate place to return it to the terrestrial world.

This is the architectural importance of the inter-satellite link. It does not merely connect two spacecraft. When deployed across hundreds or thousands of satellites, it turns a constellation into a moving network in space.

From a Collection of Satellites to an Orbital Mesh

Consider a ship operating far from land. In a traditional gateway-dependent LEO architecture, the satellite serving that ship must also be able to see a gateway capable of carrying the traffic into terrestrial networks. If no suitable gateway is visible, the service becomes difficult even though the user-to-satellite radio link may be excellent.

One solution is to construct more gateways. But ground infrastructure comes with significant cost and complexity. Every additional site may require land, power, fiber connectivity, spectrum authorization, physical security, maintenance, and regulatory approval.

An inter-satellite mesh provides another option.

The spacecraft serving the ship can send its traffic to a neighbouring satellite. That satellite may send it to another, and the packet can travel across the constellation until it reaches a spacecraft positioned above a well-connected gateway.

The result is a major change in how the network is designed.

The satellite is no longer only a relay. It begins to function as a network node.

It must understand available links, route traffic, manage capacity, report failures, apply security policies, and coordinate with the broader network. The constellation therefore starts to resemble a carrier network whose routers happen to be moving through orbit.

That distinction is one of the defining features of modern LEO communications.

Why Optical Links Are Becoming So Important

Inter-satellite links can be created using either radio-frequency technology or optical communications.

RF cross-links are well understood and can be highly reliable. Their beams are generally wider than optical beams, which provides greater tolerance to pointing errors. RF technology also benefits from decades of satellite engineering experience.

Its limitation is that radio spectrum is a finite and regulated resource. High-capacity RF cross-links require suitable bandwidth, frequency coordination, interference management, and increasingly sophisticated antennas.

Optical communications approach the problem differently.

Instead of transmitting radio waves, an optical terminal sends information through a tightly focused laser beam. Because optical wavelengths are extremely short, even a relatively small telescope can produce an exceptionally narrow beam with very high directional gain.

This makes optical ISLs attractive for large broadband constellations. They can support high data rates while concentrating energy precisely toward the receiving satellite.

But the advantage creates the challenge.

A radio beam may tolerate a small pointing error. A narrow optical beam may not.

Two spacecraft separated by hundreds or thousands of kilometres must maintain extraordinarily precise alignment while both are travelling at orbital velocity. Small errors in spacecraft attitude, vibration, thermal distortion, ephemeris information, terminal alignment, or tracking can cause the laser beam to miss the receiving telescope completely.

So the key question is not simply whether a satellite can generate a powerful laser.

The real engineering challenge is whether it can find, acquire, and continuously track another fast-moving spacecraft with extreme accuracy.

Inside an Optical Inter-Satellite Terminal

An optical ISL terminal is therefore much more sophisticated than its visible telescope might suggest.

Incoming network traffic first enters the terminal through its data interface. The modem organizes the information into frames, adds error correction, and prepares the data for optical transmission. A laser converts the electrical information into modulated light, and optical amplifiers may increase the transmitted power.

The beam then passes through steering optics before leaving the telescope.

At the other satellite, another telescope collects a very small fraction of that energy. Optical filters reject unwanted light, while photodetectors convert the received optical signal back into electrical form. Timing recovery, demodulation, and error correction reconstruct the original data.

Behind this communications chain is another equally important system: pointing and control.

The terminal must know where the remote satellite should be, where the local spacecraft is pointing, how both satellites are moving, and how accurately the telescope is aligned. Coarse pointing systems aim the terminal toward the predicted location, while fine-steering mechanisms correct smaller residual errors.

Thermal control also matters. Optical equipment is extremely sensitive to dimensional changes. A small temperature gradient can alter alignment enough to affect link performance.

This is why an optical terminal should not be viewed as a laser attached to a satellite.

It is a combined communications, navigation, pointing, control, thermal, and software system.

The Precision Challenge: Finding Another Satellite in Space

Establishing an optical link begins before either terminal sees the other.

The satellites first use orbital information to predict their future positions and determine the expected line of sight between them. This prediction allows the terminal to point toward the area in which the other spacecraft should appear.

But predicted position is never perfect.

There can be uncertainty from orbit determination, spacecraft attitude, clock differences, manoeuvres, structural alignment, or outdated ephemeris information.

The terminal therefore enters an acquisition phase. It searches the remaining uncertainty region, detects the other spacecraft's beacon, and gradually reduces the pointing error.

Once the signal has been acquired, a fine-tracking system takes over and continuously keeps the optical beam centred on the receiving telescope.

This process is commonly described as Pointing, Acquisition and Tracking, or PAT.

PAT is not an auxiliary function. It is one of the central technologies that determines whether an optical inter-satellite network can operate reliably.

There is even another subtle effect to consider.

By the time the laser beam travels from one satellite to another, the receiving satellite has moved. The transmitting spacecraft therefore has to aim slightly ahead of the receiver's current apparent position. This point-ahead correction may be very small in angular terms, but when the beam itself is extremely narrow, even tiny errors become important.

Optical communications therefore turn orbital knowledge into part of the communications link budget.

Building the Mesh: Intra-Plane and Inter-Plane Links

A constellation is normally divided into several orbital planes, with multiple satellites distributed along each plane.

The simplest inter-satellite connection is between satellites travelling in the same orbital plane. Because their relative spacing and velocity remain comparatively stable, these intra-plane links are usually easier to maintain.

A satellite might maintain one link with the spacecraft ahead of it and another with the spacecraft behind it. When repeated around an orbital plane, these links create something similar to a ring network.

The architecture is stable and relatively straightforward, but it has an important limitation: traffic has few options for moving sideways across the constellation.

That is where inter-plane links become important.

By connecting satellites in adjacent orbital planes, traffic can move not only forward and backward but also across the constellation. The network begins to resemble a grid rather than a collection of independent rings.

This greatly improves route diversity and resilience.

If one link becomes unavailable, the network may be able to redirect the traffic through another orbital plane. If one gateway becomes congested or unreachable, packets can move laterally through the constellation toward a different exit.

But cross-plane links are more demanding. Relative geometry can change rapidly, particularly near orbital seams where satellites in neighbouring planes may travel in very different directions.

The constellation therefore faces a recurring engineering trade-off: greater connectivity provides more routing options, but it also requires more terminals, more power, more tracking capability, more processing, and more sophisticated control software.

Routing Through a Network That Never Stops Moving

Perhaps the most interesting challenge appears after the optical hardware is working.

The network still needs to decide where every packet should go.

A terrestrial fiber router normally operates in a relatively stable environment. Physical links may remain unchanged for months or years unless there is a fault.

LEO routing is fundamentally different.

The topology changes as part of normal operation.

A satellite that is an excellent neighbour now may no longer be visible several minutes later. Another cross-link may soon appear. A gateway may enter or leave a satellite's coverage. Distances between nodes change continuously.

Yet orbital networks have one major advantage over many other mobile networks: their motion is highly predictable.

The position of a LEO spacecraft can be forecast accurately from orbital ephemeris. The network can therefore calculate future contact opportunities and prepare routes before existing links disappear.

Instead of waiting for a connection to fail and then reacting, the network can behave proactively.

It may install a new route, begin forwarding traffic through it, and then release the old route before service is interrupted.

This is essentially a make-before-break strategy at the network layer.

However, prediction alone is not enough.

A scheduled link may fail to acquire. A terminal may overheat. A gateway may lose power. A cross-link may become congested. A satellite may enter a safe mode.

The network therefore needs both prediction and measurement.

Orbital information tells the system what should happen.

Real-time telemetry tells it what is actually happening.

Modern LEO routing must combine both.

Why the Nearest Gateway May Be the Wrong Gateway

One of the most powerful consequences of inter-satellite networking is that the constellation can choose where traffic should return to Earth.

Without ISLs, the answer is often dictated by geometry: use whichever gateway the satellite can see.

With an orbital mesh, the network has options.

Suppose a user in one region connects to a satellite and the closest gateway is only one hop away. At first glance, that looks like the obvious path.

But that gateway may connect to a poor terrestrial network. Its fiber route may pass through a distant hub before reaching the destination application.

Meanwhile, another gateway several satellite hops away may connect directly to the cloud region or internet exchange where the customer's traffic is actually going.

In that case, keeping the packet in space slightly longer could reduce the total end-to-end journey.

This illustrates an important principle.

The best route is not necessarily the route with the fewest satellites.

It is the route that provides the best complete network path.

A sophisticated routing system may therefore consider propagation delay, congestion, available bandwidth, link reliability, gateway backhaul, terrestrial peering, security requirements, regulation, and customer priority before selecting an exit.

Do Laser Links Really Reduce Latency?

Optical inter-satellite links are sometimes associated directly with lower latency. The reality is more nuanced.

Laser light does not travel meaningfully faster than radio waves through space. Both propagate essentially at the speed of light in vacuum.

The potential advantage comes from the route itself.

Signals travelling through terrestrial fiber move more slowly because light travels more slowly through glass than through vacuum. Fiber routes are also constrained by geography, existing cable systems, landing points, and network peering arrangements.

An orbital path can sometimes take a more direct geographic route.

This can be particularly valuable across oceans or between regions where terrestrial networks take long indirect paths.

But an orbital path is not automatically better.

Every satellite hop may introduce switching, processing, and queuing delays. A badly chosen space route can therefore be slower than a well-designed terrestrial path.

Low latency is an end-to-end property.

The orbital mesh simply gives the network another route to choose from.

Extending Coverage Beyond Ground Infrastructure

One of the clearest advantages of ISLs appears where terrestrial infrastructure is weakest.

Aircraft crossing oceans, ships operating far from coastlines, polar missions, islands, deserts, remote communities, and disaster zones may all be served by satellites that have limited or no immediate access to local gateways.

An orbital mesh allows that traffic to travel toward infrastructure located elsewhere.

This can significantly reduce the need to place gateways everywhere service is required.

But it does not remove regulatory or sovereignty considerations.

Traffic may still be subject to licensing, landing rights, national security rules, data-residency requirements, or restrictions on where the network is permitted to route information.

As satellite constellations increasingly behave like global telecommunications backbones, many of the policy questions once associated mainly with international fiber networks will also apply to networks in space.

Resilience May Become Just as Important as Capacity

Inter-satellite links are often promoted for bandwidth and latency, but resilience may prove equally important.

Imagine that a major gateway loses terrestrial connectivity because of a fiber cut. A conventional gateway-dependent system might lose service across a large region.

A properly designed orbital mesh could route around the affected gateway and send traffic toward another exit.

The same principle applies if a satellite or cross-link fails.

The network can potentially bypass the failed node.

However, real resilience requires more than drawing multiple lines on a network diagram.

Two orbital routes may look different but still depend on the same satellite processor. Two gateways may appear independent while both use the same terrestrial fiber corridor. Several paths may rely on the same network controller or power facility.

True resilience therefore requires engineers to understand failure domains, not merely path diagrams.

For government, defense, aviation, enterprise, and critical infrastructure networks, this distinction can be crucial.

The Engineering Trade-Off Behind the Orbital Internet

Every improvement in an inter-satellite network has a price.

Adding more optical terminals increases route diversity, but also adds mass, power consumption, thermal load, integration complexity, and manufacturing cost.

Increasing transmission range can reduce the number of hops, but requires greater optical margin and tighter pointing accuracy.

Increasing data rate demands not only better optical terminals but also faster onboard switches, larger buffers, stronger processors, and higher-capacity internal buses.

This leads to an important engineering lesson.

A headline optical-link rate is not the same as useful network capacity.

A 100 or 200 Gbps optical terminal provides little value if the satellite cannot switch the traffic efficiently, if the link is unavailable much of the time, or if the downstream gateway lacks sufficient capacity.

The meaningful measure is therefore not simply the maximum link rate.

It is how much usable routed traffic the complete constellation can deliver reliably.

The Satellite Is Becoming a Router

Inter-satellite links represent one of the most important architectural changes in satellite communications.

They allow a LEO constellation to evolve beyond the traditional model of spacecraft acting mainly as independent relays between users and gateways.

With optical terminals, onboard packet switching, dynamic routing, predictive topology management, and intelligent gateway selection, satellites increasingly behave as nodes inside a global communications network.

The change is profound.

The network is no longer located only on Earth with satellites attached to it.

Part of the network itself is moving into orbit.

Fiber networks, mobile cores, cloud platforms, and terrestrial gateways will remain essential. But the addition of an orbital mesh introduces an entirely new communications layer capable of extending coverage, creating alternate routes, improving resilience, and potentially shortening selected global paths.

The significance of inter-satellite links is therefore much larger than the laser beam connecting two spacecraft.

They are helping transform LEO constellations from collections of satellites into something far more powerful:

a programmable global network in space.

Source: Adapted from Chapter 2, Inter-Satellite Links, in LEO Revolution — Second Edition: The Four Pillars of LEO Satellite Communications, by Karim Aziz. The book identifies inter-satellite links as one of four core technologies behind modern LEO communications. LEO_Revolution_Second_Edition_F…

Why Capacity Utilization—not Satellite Count—Determines Financial Success

Low Earth orbit satellite constellations are often discussed in terms of spacecraft numbers, launch schedules, coverage maps and technical performance. Yet behind every constellation is a much harder question:

Can the network convert billions of dollars of infrastructure into enough profitable, recurring traffic to sustain itself?

That question sits at the center of the commercial analysis in LEO Revolution. The book separates the technical foundations from the commercial and strategic foundations, with Chapter 16 specifically examining constellation capital expenditure, manufacturing, launch, ground infrastructure, terminals, operating costs, replenishment, revenue, customer acquisition, capacity utilization and return economics. LEO_Revolution_Fixed_Numbered_P…

The fundamental economic challenge is straightforward: LEO requires substantial investment before meaningful revenue appears, while satellites themselves have finite operating lives. A constellation therefore cannot be judged only by whether it works technically. It must repeatedly generate enough cash to operate the network, acquire customers and eventually replace the spacecraft already in orbit.

From CAPEX to Productive Capacity

A LEO constellation begins by consuming capital.

Money is required for research, engineering, prototypes, manufacturing facilities, satellite production, launch contracts, gateways, network-control systems, software, regulatory approvals, spectrum coordination and often customer terminals.

This means constellation capital expenditure is much broader than simply:

Number of satellites × cost per satellite.

A serious economic model must examine the entire system required to turn satellites into a functioning telecommunications service.

Timing is equally important.

Large amounts of money may be committed months or even years before the corresponding satellite begins generating revenue. Launch contracts may already be signed, factories may already be operating and employees must continue to be paid even when customer acquisition progresses more slowly than expected.

This creates one of the fundamental tensions in LEO economics:

Deploy too slowly and the network may lack sufficient coverage and capacity. Deploy too aggressively and enormous capital may be committed before demand has been proven.

Phased deployment can reduce this risk by connecting new deployment blocks to specific coverage, capacity and revenue milestones.

Manufacturing Scale Changes the Cost Equation

Traditional satellites have often been manufactured almost like specialized engineering projects.

Large LEO constellations attempt something different: industrialized spacecraft production.

Instead of producing a handful of satellites, a constellation operator may manufacture hundreds or thousands of similar spacecraft. Standardized designs, automated testing, common components, supplier agreements and repeated production can reduce the cost of each additional satellite.

But scale does not automatically produce low costs.

Production problems can erase expected savings. Rework, qualification failures, low manufacturing yield, design changes or defects affecting an entire satellite batch can significantly increase the actual cost of deployed capacity.

The meaningful economic metric is therefore not simply the advertised manufacturing cost of one satellite.

It is closer to:

Cost of producing and commissioning reliable capacity in orbit.

That distinction matters enormously. A cheap satellite that fails early may ultimately be more expensive than a higher-cost spacecraft that operates reliably for its planned lifetime.

Launch Cost Is More Than the Rocket Price

Launch economics receive enormous attention because lower launch prices helped make large LEO constellations possible.

But the advertised price of a launch vehicle tells only part of the story.

The real cost includes:

• launch vehicle and mission services,
• integration and satellite dispensers,
• transportation and range services,
• insurance and licensing,
• launch campaign operations,
• orbit raising,
• commissioning,
• and the cost of delayed revenue while satellites move into their operational positions.

Large batches can reduce the launch cost per satellite, but they also concentrate risk. A single launch failure could remove a significant portion of a deployment phase.

Smaller launches provide more flexibility and can target particular orbital planes, but the cost per spacecraft may be higher.

Therefore, launch economics should be evaluated alongside manufacturing capacity, satellite availability, replenishment requirements and the value of getting capacity into commercial service quickly.

The Ground Network Is Part of the Constellation

Satellites receive most of the public attention, but orbital capacity has little commercial value unless traffic can reach terrestrial networks and customers.

Gateways require land, antennas, RF equipment, modems, network-control systems, power, backup systems, cybersecurity and reliable terrestrial connectivity.

Fiber backhaul, data centers, cloud platforms and national network interfaces add further costs.

Operators therefore face another economic choice:

Own the ground infrastructure or buy it as a managed service?

Ownership provides greater control and may become economical at high utilization. Managed infrastructure lowers initial capital requirements but introduces recurring fees and dependence on third parties.

The correct comparison is not simply the purchase price of a gateway antenna. It is the lifecycle cost per unit of usable network capacity.

The book's executive summary similarly emphasizes that constellation economics must be evaluated as an end-to-end system incorporating satellites, ground infrastructure, customer equipment and recurring replenishment rather than as an isolated spacecraft program. LEO_Revolution_Fixed_Numbered_P…

User-Terminal Economics Can Decide the Market

One of the most underestimated costs in satellite broadband is the customer terminal.

The terminal may include the antenna, RF electronics, modem, router, enclosure and power supply, but the commercial cost continues beyond the hardware.

There may also be:

• shipping,
• import duties,
• installation,
• certification,
• warranty,
• returns,
• software support,
• replacement,
• and customer service.

For consumer markets, operators typically try aggressively to reduce terminal prices because expensive equipment creates a barrier to adoption.

That introduces another important decision: Who pays for the terminal?

The customer can purchase it outright, lease it or receive some form of subsidy.

Subsidies accelerate adoption, but they consume cash. The operator must recover that investment through future customer contribution.

The economic question is therefore not simply whether subscriber numbers are increasing.

It is:

How long does it take the contribution from each customer to recover the cost of acquiring and activating that customer?

Subscriber growth that continuously consumes more cash than customers eventually contribute is not sustainable growth.

Operating a LEO Network Is a Permanent Expense

Once satellites are launched, spending does not stop.

Operating expenditure includes satellite operations, network operations, gateways, fiber, cloud services, software, spectrum fees, insurance, cybersecurity, customer support and workforce costs.

Automation becomes particularly important at constellation scale.

A network containing thousands of spacecraft cannot depend on technicians manually managing every routine event. Fleet monitoring, configuration, customer activation, network optimization and many other activities must increasingly be automated.

But automation only reduces costs when it works reliably.

Poorly designed software may simply transfer costs from operations to engineering, outages and customer support.

The economic objective is therefore not merely automation—it is reliable automation at scale.

Replacement Is Not Optional

This is one of the largest differences between LEO and many traditional communications infrastructure projects.

LEO satellites have finite lives.

Eventually they must be replaced.

A simplified way of thinking about steady-state replacement is:

Annual replacement requirement ≈ Number of operating satellites ÷ Average satellite life

A constellation of thousands of satellites therefore develops a continuing manufacturing and launch requirement even if its geographical coverage stops expanding.

Replacement should consequently be treated as recurring capital expenditure, not as an unexpected future cost.

This changes how profitability must be viewed.

A constellation that appears profitable during its first generation of satellites but cannot generate enough cash to fund the next generation has not yet demonstrated sustainable economics.

Revenue per User Is Only Part of the Story

Average Revenue Per User—ARPU—is frequently used when discussing telecommunications economics.

But LEO networks may serve very different markets:

consumer broadband, enterprise, aviation, maritime, government, IoT and mobile-network partnerships.

Revenue therefore may need to be measured per:

• subscriber,
• site,
• vessel,
• aircraft,
• terminal,
• device,
• or unit of contracted capacity.

A high-paying customer is not automatically more valuable.

If that customer consumes enormous capacity in an already congested beam, their economic contribution can potentially be smaller than several moderate-paying customers using otherwise idle capacity.

This leads to perhaps the most important metric in LEO economics.

Capacity Utilization: The Economic Heart of LEO

A satellite can be fully operational while much of its capacity generates no revenue.

This happens because satellite capacity is geographically distributed.

A satellite may pass over:

• oceans,
• deserts,
• sparsely populated territories,
• markets where regulatory approval has not been obtained,
• or locations where few customer terminals exist.

The capacity exists technically, but economically it may be idle.

Unlike physical inventory, unused satellite capacity cannot normally be stored and sold tomorrow.

An unused beam-hour is gone forever.

This makes LEO capacity a perishable economic resource.

At the same time, average utilization can be misleading. Some urban or high-demand cells may be congested while enormous amounts of capacity remain unused elsewhere.

The real challenge is therefore not simply achieving maximum constellation utilization.

It is achieving:

Profitable utilization in the right locations, at the right times, at sustainable prices.

The book explicitly identifies capacity utilization as the bridge between engineering capability and financial performance: a constellation may provide large theoretical capacity, but commercial success depends on placing usable capacity over paying customers, at appropriate prices and with acquisition, installation and support costs under control. LEO_Revolution_Fixed_Numbered_P…

Pricing and Capacity Must Be Designed Together

Operators can improve utilization through several strategies.

Wholesale contracts with telecom operators can provide predictable demand.

Aviation and maritime services can monetize capacity across oceans and remote routes.

Enterprise and government contracts may provide higher-value commitments.

Flexible beam allocation can move resources toward areas of demand.

Different service tiers can allocate network resources according to willingness to pay and service requirements.

However, simply filling the network is not enough.

Selling enormous volumes of capacity at extremely low prices may improve utilization while weakening cash generation.

Conversely, protecting high prices while leaving most capacity idle will not finance a constellation either.

The objective is therefore neither the highest price nor the highest utilization.

It is the combination of price, utilization and margin that maximizes sustainable economic contribution.

Break-Even Must Include the Whole System

Statements such as "the constellation has reached break-even" can mean very different things.

A company might mean that customer revenue exceeds direct customer costs.

Another might mean operating revenue covers network operating expenses.

A more demanding definition would include financing costs, replacement satellites and continuing infrastructure investment.

For a LEO constellation, the strongest measure is not simply whether current revenue exceeds current operating expenses.

It is whether the business can fund:

operations + customer growth + network maintenance + satellite replenishment + financing obligations

while still producing an acceptable return on invested capital.

That is a much higher hurdle.

The Real Economic Equation

The economics of LEO can ultimately be reduced to a simple relationship:

Investment → Capacity → Utilization → Revenue → Margin → Reinvestment

Manufacturing scale can lower satellite costs.

Reusable launch systems can reduce deployment costs.

Phased-array terminals can become cheaper.

Automation can reduce operating expenses.

But none of these developments guarantees financial success.

A constellation becomes economically sustainable only when enough of the capacity created by those investments is repeatedly purchased by customers at prices that cover the full lifecycle cost of the network.

That is why the future LEO market will not be determined simply by who launches the most satellites.

It will increasingly be determined by who monetizes orbital capacity most efficiently.

And that may ultimately be the most important economic lesson of the LEO revolution.

Based on Chapter 16, "The Economics of LEO Constellations," from LEO Revolution by Karim Aziz. .

By Karim Aziz

Low Earth orbit satellite networks promise global coverage, lower latency, and broadband connectivity for people, vehicles, aircraft, ships, and connected machines. However, these benefits cannot be delivered by the orbit alone. They depend heavily on one of the most important technologies in the entire system: the phased-array antenna.

Unlike a traditional parabolic dish, a phased-array antenna can direct its radio beam electronically without physically moving the antenna. This ability allows a terminal to track fast-moving LEO satellites, switch between them, manage multiple beams, and maintain communication as the constellation moves across the sky.

Phased arrays are therefore much more than improved antennas. They are one of the four technological pillars that make modern LEO satellite communications practical.

From Mechanical Movement to Electronic Steering

Traditional satellite communications were designed mainly for geostationary satellites. A GEO satellite appears fixed in the sky because it orbits at the same rotational rate as Earth. Once a parabolic antenna is correctly aligned, it can remain pointed in the same direction.

LEO satellites operate very differently. They travel at approximately 7.5 kilometres per second and cross the visible sky within minutes. A user terminal cannot remain pointed at one fixed orbital position. It must continuously follow one satellite and then transfer the connection to another.

Mechanical tracking antennas can perform this function, but they contain motors and moving components that add weight, maintenance requirements, noise, and potential points of failure. They are also unsuitable for many mass-market applications.

A phased-array antenna replaces mechanical movement with electronic control. It contains many small antenna elements arranged across a flat surface. Each element transmits or receives the same radio signal with a carefully controlled difference in timing, known as phase.

When the signals from these elements combine, they reinforce one another in a selected direction and partially cancel in other directions. By changing the phase applied to each element, the antenna can steer its beam almost instantly.

The antenna itself does not rotate. The beam moves electronically.

Why Phased Arrays Are Essential for LEO

A LEO user terminal may see several satellites during the same period. As one satellite approaches the horizon, another may be moving into a better position. The terminal must determine which satellite offers the most suitable link and manage the transition without interrupting the service.

Phased-array antennas support this process by providing:

• Rapid electronic beam steering

• Continuous tracking of moving satellites

• Fast handover between satellites

• Multiple simultaneous communication beams

• High antenna gain from a relatively flat surface

• Improved resistance to unwanted interference

These capabilities make it possible to install satellite terminals on homes, vehicles, aircraft, ships, trains, and industrial equipment.

For mobile platforms, the antenna may also need to compensate for the movement of the platform itself. An aircraft changes its direction, altitude, pitch, and roll while the satellite continues moving across the sky. The antenna-control system must calculate both movements and continuously adjust the radio beam.

This is one reason electronically steered antennas have become central to aviation and maritime connectivity.

How Beamforming Works

The total signal produced by a phased array depends on the relationship between its individual antenna elements.

If all elements transmit with the correct phase relationship, their energy combines to produce a strong beam in the required direction. If the phase pattern changes, the beam is redirected.

This process is known as beamforming.

Three main beamforming architectures are commonly used.

Analog beamforming controls the signals through radio-frequency phase shifters. It is relatively power-efficient and economical, but it normally offers limited flexibility and may support only one or a small number of beams.

Digital beamforming converts the signals from many antenna elements into digital form. A processor can then create and control multiple beams, suppress interference, and allocate capacity dynamically. Digital beamforming offers greater flexibility but requires more processing power, electrical power, cooling, and system complexity.

Hybrid beamforming combines analog and digital techniques. It provides more flexibility than a purely analog design while reducing the processing and power requirements of a fully digital architecture.

The best approach depends on the application. A residential broadband terminal, a satellite payload, an aircraft antenna, and a direct-to-device system do not have the same performance, size, power, or cost requirements.

Two Sides of the Satellite Link

Phased arrays are important at both ends of a LEO connection.

On the satellite, large antenna arrays create multiple spot beams over the service area. These beams allow the operator to concentrate radio power where it is needed, reuse frequencies in different geographic areas, and distribute capacity according to demand.

The satellite can adjust beam coverage as it moves over cities, rural regions, oceans, aviation routes, or areas affected by emergencies. Future systems may use increasingly flexible digital payloads to create, resize, redirect, or deactivate beams through software.

At the user end, the phased array must locate and track the satellite while maintaining sufficient signal strength. It must also manage polarization, frequency, Doppler shift, interference, and handover.

These functions must often be performed inside a compact and affordable terminal. This creates one of the most difficult commercial challenges in LEO broadband: delivering high antenna performance without making the terminal too expensive for the intended customer.

The Importance of Aperture and Frequency

The physical size of an antenna affects its gain and beamwidth. A larger aperture can normally produce a narrower and stronger beam. However, the required size also depends on the operating frequency.

At higher frequencies, such as Ku-band and Ka-band, antenna elements can be smaller and placed closer together. This makes it possible to integrate many elements into a relatively compact panel.

Higher frequencies also create engineering challenges. Signals can be affected more strongly by rain, atmospheric conditions, manufacturing tolerances, and losses in electronic components. Ka-band systems may offer greater bandwidth, but they require accurate design and effective link-margin management.

Lower-frequency systems provide different advantages. They may experience less rain attenuation and can be more suitable for certain mobile or direct-to-device services. However, their longer wavelengths require larger antenna elements and greater spacing between them, making a high-gain array physically larger.

The frequency decision therefore influences the size, cost, performance, power consumption, and intended service of the antenna.

Scan Loss and Field of View

A phased array normally provides its strongest performance when the beam points directly away from the antenna surface. As the beam is steered toward a lower elevation angle, its effective aperture becomes smaller and its gain decreases.

This effect is known as scan loss.

The antenna may also experience increased sidelobes, polarization degradation, and greater sensitivity to element-spacing errors at wide scan angles. As a result, an antenna advertised as supporting a wide field of view may not provide equal performance in every direction.

For a LEO terminal, this creates an important design trade-off. Maintaining a link to a satellite close to the horizon increases coverage time, but the longer slant range and antenna scan loss reduce the available link margin.

Operators may therefore establish a minimum operating elevation angle. Below that angle, the terminal may switch to another satellite or wait until a better link becomes available.

Handover Without Service Interruption

LEO connectivity depends on frequent handovers. A user terminal may change satellites several times during a single session.

An advanced phased array can form two beams simultaneously. One beam maintains the active connection while the second searches for or establishes communication with the next satellite. This approach supports a make-before-break handover, reducing the possibility of service interruption.

The antenna does not make the decision alone. Handover requires coordination among the terminal, satellite, gateway, network-management system, and routing platform.

Orbital information allows the network to predict when a satellite will enter or leave the terminal’s field of view. The network can prepare the next connection before the existing link becomes weak.

Successful handover is therefore a system-level function in which the antenna provides the physical beam-steering capability.

The Cost and Power Challenge

Phased arrays offer remarkable performance, but they are not simple or inexpensive devices.

A broadband terminal may contain hundreds or thousands of antenna elements, phase-control circuits, amplifiers, filters, frequency converters, control processors, and thermal-management components. Small manufacturing differences between elements can affect the overall beam pattern.

Power consumption is another major issue. Electronic components generate heat, and excessive temperature can reduce performance and reliability. This is particularly important for terminals installed on aircraft, vehicles, rooftops, or equipment operating in hot environments.

The design must balance several competing objectives:

• Antenna gain

• Beam-steering range

• Number of simultaneous beams

• Data throughput

• Power consumption

• Thermal performance

• Physical size and weight

• Manufacturing cost

• Reliability and service life

An antenna can perform extremely well technically and still fail commercially if it is too expensive, consumes too much power, or cannot be manufactured at scale.

Phased Arrays and Direct-to-Device Connectivity

Direct-to-device connectivity places even greater responsibility on the satellite antenna.

An ordinary smartphone has a small antenna and limited transmission power. It cannot provide the same link performance as a dedicated broadband satellite terminal. The satellite must compensate by using a large receive array, high antenna gain, precise beamforming, sensitive receivers, and advanced signal processing.

This is why large deployable antennas and massive phased arrays are becoming important in direct-to-cell systems. They allow the satellite to create narrow beams and detect weak signals transmitted from standard or lightly modified mobile devices.

However, a larger array does not create unlimited capacity. Every beam must share the spacecraft’s available spectrum, power, processing capability, and feeder-link resources.

The phased array makes the connection possible, but the complete network determines how many users can be served and what services can be delivered.

A Strategic Technology, Not Just an Antenna

Phased-array technology affects almost every part of a LEO satellite business.

It influences satellite design, launch mass, user-terminal pricing, coverage, capacity, service quality, energy consumption, manufacturing scale, and the markets that the operator can serve.

For consumers, success means a reliable and affordable terminal that requires little technical knowledge. For aviation and maritime users, it means maintaining connectivity while the platform and satellites are both moving. For direct-to-device systems, it means closing an extremely difficult radio link with a low-powered smartphone.

The future of LEO communications will depend partly on how quickly the industry can improve antenna performance while reducing cost, power consumption, size, and manufacturing complexity.

Conclusion

Phased-array antennas transform a fixed radio surface into an intelligent and electronically controlled connection. They allow beams to follow satellites, transfer links, concentrate capacity, reduce interference, and support moving users without mechanical tracking.

Without this technology, modern LEO networks would struggle to deliver practical broadband, mobility, and direct-to-device services at scale.

The satellite constellation may be the most visible part of a LEO network, but the phased-array antenna is what enables users to reach it. That is why it deserves its position as the first of the four pillars of modern LEO satellite communications.

www.leosatconnect.com