Chapter 1
A satellite internet service is a long chain of physical and logical parts: a constellation moving overhead at orbital speed, laser or radio links between spacecraft, gateway stations tied to the terrestrial backbone, and a small terminal on a roof, a ship or an aircraft. WIA-COMM-005 describes that whole chain with one vocabulary and one set of data shapes. This chapter introduces the standard, the physics behind it, and the records an implementer will exchange.
The standard targets a specific gap: places where fibre and cellular networks do not reach. Its stated purpose is universal access for remote, rural, maritime and aviation environments, under the principle 弘益人間 (Benefit All Humanity). To serve those users it has to cover low-, medium- and geostationary-orbit systems, because each orbit trades latency against coverage in a different way. The sections below cover how the standard bounds its scope and what the core terms mean.
The specification defines a technical framework for global broadband delivered through space-based networks. It names LEO constellations of the Starlink, OneWeb and Kuiper kind, MEO systems, and traditional GEO satellites. The scope is end to end, and it includes the following areas:
Most of the standard's arguments rest on a small group of terms, which the specification fixes as follows.
| Term | Meaning in the standard |
|---|---|
| LEO | Low Earth Orbit, 500-2000 km altitude |
| MEO | Medium Earth Orbit, 2000-35,786 km altitude |
| GEO | Geostationary Earth Orbit, 35,786 km altitude |
| ISL | Inter-Satellite Link, optical or RF, between satellites |
| EIRP | Effective Isotropic Radiated Power |
| G/T | Gain-to-noise-temperature ratio of a receiving system |
| Ka-band | 26.5-40 GHz |
| Ku-band | 12-18 GHz |
| V-band | 40-75 GHz |
The ranges for MEO overlap slightly with the later architecture section, which describes MEO systems at 8,000-20,000 km. The 2000-35,786 km range in the terminology is the definition of the tier, and the narrower range describes typical deployments.
Three physical effects shape every design decision in the standard: the geometry of the orbit, the loss of signal over distance, and the relative motion between a satellite and the ground. The two sections below treat them in turn.
A typical LEO configuration in the specification uses altitudes of 550 km (Starlink), 600 km (OneWeb) and 590 km (Kuiper). Inclinations of 53°, 70° and 87° give different coverage patterns, and the orbital period is about 90-100 minutes. A constellation spreads 20-80 planes of 20-80 satellites each, for totals from 300 to more than 40,000 satellites. Inclination decides who is served. Near-polar orbits at 87-90° cover the whole globe including the poles, which suits scientific, military and emergency use. Inclinations of 53-70° concentrate capacity on populated mid-latitudes, which suits commercial broadband.
Satellites are spaced evenly so that coverage is continuous. The spec gives the phasing rules and a worked case:
Phase Angle = 360° × (plane_number / total_planes)
Satellite Spacing = 360° / satellites_per_plane
Starlink Phase 1 example:
72 orbital planes, 22 satellites per plane
1,584 total satellites, 5° between planesThe 5° figure follows directly from dividing 360° by 72 planes. Spacing within a plane works out to about 16.4° for 22 satellites.
Whether a terminal can use a satellite depends on the minimum elevation angle. The spec relates it to altitude like this:
cos(θ) = R_e / (R_e + h)
r = R_e × arccos[R_e × cos(θ) / (R_e + h)]Here θ is the minimum elevation angle, typically 25-40°, R_e is 6,371 km, and h is the satellite altitude. The second expression gives the radius of the ground circle a satellite can serve at that elevation. A higher minimum elevation shrinks each satellite's footprint but avoids low, obstructed, rain-prone slant paths. This is why the constellation descriptor carries an expected minimum elevation alongside altitude and inclination.
The standard expresses link quality as a chain of terms that an implementer can compute and audit:
EIRP (dBW) = Tx Power (dBW) + Antenna Gain (dBi) - Losses (dB)
Path Loss (dB) = 20 log10(d) + 20 log10(f) + 92.45
SNR (dB) = EIRP - Path Loss - Noise Figure + G/T
Capacity (Mbps) = Bandwidth × log2(1 + SNR)In the path loss expression d is in kilometres and f in gigahertz. The logarithmic form means that doubling either distance or frequency adds about 6 dB. That is why Ka-band links over several hundred kilometres still demand high-gain antennas at both ends. The capacity line is the Shannon form, so the SNR inside the logarithm must be a linear ratio, not decibels. The detailed downlink budget in the API phase sets satellite EIRP at 50-60 dBW, free-space loss at 165-175 dB, atmospheric absorption at 0.5-2 dB, and rain fade at 1-10 dB. It gives typical SNR of 15-25 dB in clear sky, falling to 5-15 dB in rain. Rain fade is the main reason capacity varies from hour to hour at Ka-band and above.
Relative motion creates Doppler shift, which the terminal modem must track:
Δf = (v / c) × f0
v ≈ 7.5 km/s (LEO satellite relative to ground)
c = 299,792 km/sApplying this to a 28 GHz carrier gives a shift on the order of 700 kHz at the extreme. A modem that cannot pre-compensate for that offset cannot lock onto the carrier at all.
Motion also forces handover. The spec predicts the time a satellite stays usable with:
t_handover = arccos(R_e / (R_e + h)) / ωHere ω is the satellite's angular velocity. A low-altitude satellite crosses the sky quickly. A terminal must therefore continuously prepare its next serving satellite, and the standard names "seamless satellite-to-satellite transitions" as a handover requirement.
The standard does not publish pricing or subscriber data, so this section describes how the three orbit tiers divide the market. That division follows from physics. The comparison below is assembled from the specification's own figures.
| Tier | Altitude | Period | Round-trip latency | Typical role |
|---|---|---|---|---|
| LEO | 500-2000 km | about 90-100 min | 20-40 ms | Broadband, real-time applications |
| MEO | 8,000-20,000 km typical | 6-12 hours | 100-150 ms | Navigation, regional fill, LEO backup |
| GEO | 35,786 km | 23 h 56 min | 500-600 ms | TV broadcast and DTH, legacy maritime and aviation, weather, emergency backup |
The latency column explains most of the competitive dynamics. A GEO signal must climb 35,786 km and return, and that distance alone costs roughly a quarter of a second in each direction. This is why GEO remains strong for broadcast, where one transmission reaches many receivers and delay does not matter, but is a poor fit for interactive traffic. LEO cuts the propagation delay to a few milliseconds. The price is scale: a LEO service needs hundreds to tens of thousands of satellites, plus a global network of gateways, before coverage is continuous.
Cost is therefore driven by constellation size, gateway count (10-50 gateways globally per the spec), and terminal hardware. Terminal power is a practical constraint on adoption. The spec lists 20-50 W idle, 50-150 W active, and 100-200 W peak. A vessel or an off-grid household has to budget for that draw. The standard acknowledges that no single orbit wins everywhere and describes a hybrid pattern:
The use cases named in the standard fall into three groups: people with no terrestrial option, mobile platforms, and specialised services.
Inter-satellite links matter most in remote and oceanic service. With optical ISLs at 1550 nm carrying 1-100 Gbps over distances up to 5,000 km, traffic can cross the constellation in space and descend at a gateway near an internet exchange point. The spec's mesh section lists reduced ground-station dependency, lower latency through space routing, greater resilience, and global coverage without ground infrastructure. Routing options include OSPF-based routing, BGP between constellations, and Delay-Tolerant Networking.
The ground segment determines whether those advantages reach the customer. A gateway uses a 3-13 m antenna at Ka-band, 27.5-30 GHz uplink and 17.7-20.2 GHz downlink, and sustains 10-100 Gbps. Its teleport combines high-power and low-noise amplifiers, frequency converters, DVB-S2X or 5G NR baseband, LDPC or turbo error correction, and a fibre backbone with CDN peering. Placement guidance favours sites near internet exchange points with low latency to major data centres, geographic diversity, and a regulatory-friendly jurisdiction.
Satellite systems are regulated before they are engineered. The standard lists ITU Radio Regulations and orbital debris mitigation as its compliance anchors. Frequency coordination matters because the bands overlap in use. Ka-band (26.5-40 GHz) carries high-capacity user links, Ku-band (12-18 GHz) serves legacy and backup systems, V-band (40-75 GHz) is for future high-throughput use, and Q/V-band (40-50 GHz) is listed for gateway feeder links. Optical ISLs have a regulatory advantage the spec calls out, since they need no frequency licensing. The CLI includes a debris check that takes an altitude and an inclination, which reflects that debris compliance is evaluated against the orbital shell.
On the data side, every Phase 1 record uses UTF-8 JSON with RFC 8785 canonicalisation, Ed25519 signatures and ULID identifiers. Orbital elements use the legacy Two-Line Element format for compatibility, plus CCSDS 502.0-B Orbital Ephemeris Messages for high-precision propagation. Within the 1.x line, hosts may add optional fields but may not remove existing ones.
An implementer faces two kinds of work: getting the RF and orbital numbers right, and moving the descriptive records between parties in a verifiable form. The terminal side deserves attention first. The spec describes phased arrays of 30-60 cm (Starlink) or 50-70 cm (OneWeb) with 1,000-2,000 elements, electronic steering to ±60° from zenith, and dual circular or linear polarisation. Steering follows θ = arcsin(λ × Δφ / 2πd), where Δφ is the phase shift between elements and d the element spacing. The modem uses DVB-S2X modulations from QPSK to 32APSK, LDPC coding with rates from 1/4 to 9/10, and symbol rates of 100-500 Msps. Typical user rates are 100-500 Mbps down and 20-50 Mbps up.
The constellation descriptor is the main Phase 1 record. The worked example in the specification looks like this, shortened before its signature field:
{
"wia_satinternet_version": "1.0.0",
"type": "constellation_descriptor",
"constellation_id": "starlink-shell-1",
"tier": "leo",
"satellite_count": 1584,
"altitude_km": 550,
"inclination_deg": 53.0,
"expected_minimum_elevation_deg": 25.0,
"isl_capable": true,
"first_launch_at": "2019-11-11T12:56:00Z",
"operator_id": "did:wia:satinternet:operator-A"
}The tier field takes leo, meo or geo. The other record families are described in prose. An inter-satellite link record carries the link endpoints, the optical wavelength, the link rate (typically 100 Gbps for laser ISLs) and a link quality estimate, and the satellite operator signs it so consumers can verify which constellation produced the link. Ground-station records carry station identity, geographic location, supported frequency bands and operational state. User-terminal records carry terminal identity, owner, supported bands and the last-known link metrics used for performance monitoring.
The TypeScript SDK turns the formulas of section 1.3 into calls. The following excerpt is taken from the spec's own example:
import { SatelliteNetwork, calculateLinkBudget, predictHandover } from '@wia/comm-005';
const network = new SatelliteNetwork({
constellation: 'LEO', altitude: 550, inclination: 53,
totalSatellites: 1584, orbitalPlanes: 72
});
const linkBudget = calculateLinkBudget({
distance: 600, frequency: 28, txPower: 10, antennaGain: 35, bandwidth: 250
});
const handover = predictHandover({
currentSatelliteId: 'SAT-1234',
userLatitude: 37.7749, userLongitude: -122.4194, elevationAngle: 25
});The command-line tool exposes the same calculations for quick checks, for example wia-comm-005 doppler --velocity 7500 --frequency 28000 or wia-comm-005 beam-steer --users 1000 --coverage-area 500. A sensible first step is to reproduce the SDK's link budget by hand, using the path loss formula above, and compare it with the tool's output. That confirms unit handling for kilometres, gigahertz, megahertz and watts before any live data is involved.
The specification points in a few directions. V-band is described as the future high-throughput band, with bandwidth up to 10 GHz. That is far more than the 3.5 GHz listed for Ka-band, but V-band suffers more from rain fade, so link-budget margin and gateway diversity will matter even more. Optical ISLs already span 1-100 Gbps, and their reliance on pointing means beam divergence of under 10 microradians and a pointing-loss term in every budget. Constellation sizes up to 40,000 satellites imply a heavier load on visibility prediction, handover scheduling and debris compliance. Hybrid architectures should deepen as LEO, MEO and GEO each take the role they do best. Delay-Tolerant Networking is listed for deep-space routing and shows the standard's reach beyond terrestrial broadband.
WIA-COMM-005 treats satellite internet as one system built from orbital geometry, link physics, ground infrastructure and signed records. Altitude, inclination and phasing decide coverage. The minimum elevation angle decides which satellites a terminal can use. Free-space loss, rain fade and Doppler shift decide whether the link closes, and the handover formula decides how long it stays closed. The three orbit tiers differ chiefly in latency, which is why the standard favours hybrid designs: LEO for interactive access, MEO for regional fill, and GEO for broadcast and backup. The constellation descriptor, the ISL record, and the ground-station and terminal records make these properties machine-readable.
The next chapter takes up the orbital layer in detail. It starts with LEO shells, where the plane counts, phase angles and elevation geometry introduced here are used to size a constellation and predict what a terminal at a given latitude will see.