Module 1: Satellite Networking Foundations

Core building blocks: what a satellite network is, orbit mechanics, and spectrum planning.

What a Satellite Network Is

A satellite network links users to services using space and ground assets together.

The User Segment

The user segment encompasses every device that connects end users to the satellite network. Traditional VSAT (Very Small Aperture Terminal) dishes range from 0.75 m to 2.4 m in diameter and are widely deployed in maritime, enterprise, and rural broadband scenarios. Hughes Jupiter terminals, for example, deliver up to 100 Mbps to fixed-site consumers across North America. At the smaller end, handheld satellite phones from Iridium and Thuraya serve voice and low-data needs, while NTN-capable smartphones from Qualcomm and MediaTek chipsets are beginning to bridge the gap between terrestrial and space-based connectivity. IoT terminals — often solar-powered with omnidirectional antennas — report telemetry from oil rigs, cargo ships, and environmental sensors in locations no cell tower could reach.

As of early 2025, SpaceX has launched over 6,000 Starlink satellites and shipped more than 4 million user terminals worldwide, making it the largest commercial satellite broadband constellation in history.

The Space Segment

The space segment includes the satellites themselves and the orbital infrastructure supporting them. Each satellite consists of two primary subsystems: the bus (power, propulsion, thermal control, attitude determination) and the payload (transponders, antennas, and onboard processors that handle communications). A typical GEO communications satellite carries 24-72 transponders, each capable of 36-72 MHz of bandwidth. LEO constellations like Starlink use flat-panel phased-array antennas and inter-satellite laser links (ISLs) to route traffic across the constellation without touching the ground. OneWeb operates 648 satellites in a 1,200 km orbit, while Telesat Lightspeed targets 298 satellites in a 1,015 km LEO orbit with advanced onboard processing.

Satellite Size Matters

Not all satellites are the same size — and for D2D (direct-to-device) services, size is directly related to capability. A CubeSat fits in the palm of your hand (10 cm³) and is used for basic IoT data collection. Lynk Global's D2D satellites have a 4 m² antenna — enough for text messaging to standard phones. Starlink Gen2 satellites are 25 m² and weigh 1,250 kg, enabling broadband via dedicated terminals. But AST SpaceMobile's BlueBird 6 is a different beast entirely: at 6,100 kg with a 223 m² phased array antenna (the largest commercial array ever deployed in LEO), it can focus energy into tight spot beams to deliver broadband directly to unmodified smartphones. The general rule: the bigger the satellite antenna, the more bandwidth it can deliver to smaller ground devices.

Launch Economics: Why Now?

The explosion of LEO constellations was only possible because SpaceX fundamentally changed launch economics through reusable rockets. Before Falcon 9, launching payload to LEO cost roughly $54,000 per kilogram. SpaceX's reusable Falcon 9 brought that down to approximately $2,600/kg for customers (with SpaceX's internal cost estimated around $630/kg). Looking ahead, Starship-class vehicles aim to push costs below $100/kg by 2030. This 50x cost reduction is why companies can now afford to launch thousands of satellites — Starlink alone has over 10,000 in orbit.

The Ground Segment

Gateways and teleports form the critical bridge between space and terrestrial networks. A gateway earth station typically uses large parabolic antennas (7-13 m for GEO, smaller tracking antennas for LEO) to maintain high-throughput feeder links with the satellite. Major teleport facilities — such as Hughes in Germantown, Maryland, or SES in Betzdorf, Luxembourg — house dozens of antennas, redundant fiber connections to internet exchange points, and Network Operations Centers (NOCs) that monitor satellite health, traffic loads, and link quality around the clock. Fiber backhaul from the gateway to the nearest internet peering point is essential; even a 1 Tbps satellite is bottlenecked if the gateway has only a few Gbps of terrestrial connectivity.

The ground segment often accounts for 30-40% of total satellite network capital expenditure. Poor ground infrastructure is the most common cause of user-perceived performance problems in satellite broadband.

How the Segments Interconnect

All three segments must work in concert. The user terminal acquires the satellite signal (or is assigned a beam), the satellite relays or processes the traffic, and the gateway hands it off to the terrestrial network. Monitoring, capacity planning, and fault management span all three layers. Understanding this three-segment model is foundational to every other topic in this learning path, because failures or constraints in any single segment ripple across the entire service.

Why it matters: It sets the baseline for all tabs in this site and prevents "satellite-only" thinking.

Orbit Basics: GEO, MEO, LEO

Different orbit heights trade off latency, coverage footprint, and constellation size.

Geostationary Orbit (GEO)

GEO satellites orbit at exactly 35,786 km above the equator, where their orbital period matches Earth's rotation, making them appear stationary from the ground. This fixed position eliminates the need for tracking antennas, which is why GEO dominates broadcast television (DirecTV, Sky) and wide-area VSAT services. However, the immense distance means a signal round-trip takes 550-700 ms, making GEO poorly suited for real-time interactive applications like voice calls or video conferencing without significant optimization. A single GEO satellite can illuminate roughly one-third of Earth's surface, so just three satellites spaced 120 degrees apart provide near-global coverage between approximately 70 degrees N and 70 degrees S latitude. Major GEO operators include Intelsat, SES, Eutelsat, and Viasat.

Three GEO satellites can cover nearly the entire globe, but the 550-700 ms round-trip delay makes the orbit unsuitable for latency-sensitive applications without Performance Enhancing Proxies (PEPs) and protocol optimization.

Medium Earth Orbit (MEO)

MEO spans 2,000 km to 35,786 km altitude and offers a compelling balance between coverage footprint and latency. The most well-known MEO constellation is GPS, orbiting at 20,200 km with 31 operational satellites. For broadband, SES O3b mPOWER operates at 8,062 km altitude with a constellation of high-throughput satellites, delivering round-trip latency of 120-150 ms, roughly comparable to a long terrestrial path. MEO satellites move relative to the ground but slowly enough that tracking is manageable with moderate antenna systems. Each MEO satellite covers a larger area than a LEO satellite but less than GEO, so constellation sizes are moderate, typically 12-24 satellites for regional or global coverage.

Low Earth Orbit (LEO)

LEO extends from 160 km to 2,000 km altitude. Starlink operates at approximately 550 km, OneWeb at 1,200 km, and Telesat Lightspeed at 1,015 km. The proximity to Earth yields excellent latency, typically 25-60 ms round-trip, making LEO suitable for interactive applications, gaming, and voice. The trade-off is that each satellite covers a small footprint and moves across the sky in about 5-10 minutes, requiring frequent handovers between satellites and large constellations (hundreds to thousands of satellites) for continuous coverage. LEO also introduces significant Doppler shift, up to plus or minus 24 ppm at Starlink altitudes, which user terminals and the network must continuously compensate for.

LEO satellites move at roughly 7.5 km/s and cross the visible sky in about 5-10 minutes. This speed creates Doppler frequency shifts of up to 24 parts per million that both terminals and network protocols must compensate for in real time.

Highly Elliptical Orbits (HEO)

HEO orbits like the Molniya orbit (apogee around 40,000 km, perigee around 500 km, 63.4 degree inclination) spend most of their orbital period high above high-latitude regions, providing extended dwell time over areas that GEO cannot reach effectively, such as the Arctic, northern Russia, and northern Canada. A system of 2-3 Molniya satellites can provide continuous high-latitude coverage. The SDA (Space Development Agency) defense constellations explore HEO variants for persistent polar coverage. While niche, HEO is critical for Arctic shipping routes, northern military operations, and regions above 70 degrees latitude where GEO geometry fails.

Why it matters: Orbit choice drives user experience and network design complexity.

Spectrum and Frequency Planning

Different bands are used for different performance and resilience needs.

L-Band and S-Band: Resilience and Mobility

L-band (1-2 GHz) is the workhorse of mobile satellite services. Inmarsat's global network operates primarily in L-band, providing maritime voice, safety-of-life services, and aeronautical communications. L-band signals penetrate foliage, light rain, and building walls better than any higher band, making it ideal for handheld and mobile terminals. S-band (2-4 GHz) is used for mobile satellite services and is the primary band for Direct-to-Device (D2D) NTN experiments under 3GPP, as it offers a good balance between propagation resilience and available bandwidth. Both bands are limited in total capacity due to narrow spectrum allocations, typically 30-40 MHz per operator. Crucially, L-band and S-band are the primary bands for Direct-to-Device (D2D) satellite services — they are the only satellite bands that can reach a standard smartphone without a dedicated dish or terminal. Modern chipsets like the Qualcomm Snapdragon X80 and later include native NTN support in these bands, enabling phones to connect directly to LEO satellites.

C-Band and Ku-Band: The Broadcast and VSAT Workhorses

C-band (4-8 GHz) was the original satellite communications band and remains critical for broadcast distribution and backbone connectivity, particularly in tropical regions where its weather resilience outperforms higher bands. Ku-band (12-18 GHz) is the most widely used VSAT band globally, powering maritime broadband, enterprise connectivity, and consumer services like early HughesNet. Ku-band offers moderate throughput with manageable rain fade in most climates, though tropical regions can experience 6-10 dB of attenuation during heavy downpours.

Rain fade increases dramatically with frequency: L-band experiences negligible fade, Ku-band loses 3-10 dB in heavy rain, and Ka-band can lose 15-20 dB or more, requiring adaptive coding, power control, and site diversity to maintain service.

Ka-Band and V-Band: High Throughput and the Future

Ka-band (26-40 GHz) powers the high-throughput satellite (HTS) revolution. With wide spectrum allocations (up to 3.5 GHz), Ka-band enables satellites like ViaSat-3 to deliver over 1 Tbps of total capacity. However, severe rain fade at these frequencies demands sophisticated countermeasures including adaptive coding and modulation (ACM), uplink power control, and gateway site diversity with 20-50 km separation between redundant sites. V-band (40-75 GHz) is emerging as the next frontier for feeder links, where even wider spectrum is available. Starlink and other LEO operators are exploring V-band gateway links to reduce the number of gateway sites needed.

Link Budget Basics and Frequency Reuse

A link budget accounts for every gain and loss in the signal path: transmit power, antenna gain, free-space path loss, atmospheric attenuation, and receiver sensitivity. Higher frequencies suffer greater free-space path loss (proportional to frequency squared), which must be offset by larger antennas or higher power. Modern HTS satellites overcome bandwidth limits through frequency reuse via spot beams. By dividing coverage into hundreds of narrow beams and assigning the same frequency to non-adjacent beams (typically a 4-color reuse scheme), total system capacity multiplies dramatically compared to a single wide beam using the full spectrum once.

Frequency reuse via spot beams is the single most important technique in modern satellite capacity engineering. A satellite with 200 spot beams and 4-color reuse can achieve 50x the capacity of a wide-beam satellite using the same total spectrum.

Why it matters: Band planning is central to both user experience and deployment economics.

Payloads, Beams, and Capacity

Spot beams and payload design determine how much traffic a satellite can serve.

Bent-Pipe (Transparent) Payloads

In a bent-pipe architecture, the satellite acts as a simple relay. It receives the uplink signal, shifts it to a different frequency, amplifies it, and retransmits it on the downlink. All signal processing — demodulation, decoding, routing, and re-encoding — happens on the ground. This approach is proven, reliable, and cost-effective to build. The vast majority of GEO communications satellites in orbit today use bent-pipe payloads. The trade-off is inflexibility: capacity allocation across beams is fixed at launch or limited to analog switching between predetermined configurations.

Regenerative (Onboard Processing) Payloads

Regenerative payloads demodulate and decode the signal onboard the satellite, then re-encode and retransmit it. This enables packet-level switching between beams, dynamic capacity reallocation, and even full base station (gNB) processing in orbit under 3GPP Release 19. Regenerative payloads reduce round-trip latency in multi-hop scenarios and enable inter-satellite link (ISL) routing without returning traffic to the ground. The downside is higher complexity, power consumption, and cost per satellite, plus the risk that onboard processing hardware becomes obsolete faster than the satellite's orbital lifetime.

ViaSat-3 delivers over 1 Tbps of total capacity using Ka-band spot beams — more than the entire rest of the commercial GEO fleet combined at the time of its design. Each satellite covers a full hemisphere with roughly 2,000 spot beams.

Spot Beams and Frequency Reuse

Wide-beam satellites illuminate large regions with a single beam, reusing the spectrum only once. Spot-beam satellites divide coverage into hundreds or thousands of narrow beams, each covering a small geographic area. Non-adjacent beams can reuse the same frequency, multiplying total system capacity. The standard approach uses a 4-color frequency reuse scheme, where four different frequency/polarization combinations tile the coverage area so that no two adjacent beams share the same channel. This technique is what makes high-throughput satellites (HTS) possible, boosting capacity by 20-100x over traditional wide-beam designs.

Feeder Links and User Links

The feeder link connects the satellite to its ground gateway, typically using Ka-band or V-band with very high throughput. The user link connects the satellite to end-user terminals, often in Ku-band or Ka-band. Feeder links must carry the aggregate traffic of all user beams they serve, so they require large antennas and high power. Beam hopping is an emerging technique where the satellite rapidly switches its transmit beam across different coverage cells on a time-division basis, dynamically allocating capacity to cells where demand is highest rather than distributing it evenly across all beams at all times.

Beam hopping allows a satellite to concentrate 3-5x more capacity on high-demand cells during peak hours compared to static beam allocation, dramatically improving resource efficiency for operators serving uneven traffic patterns.

Why it matters: It clarifies why capacity differs by location and time.