Medium Earth orbits
Medium Earth orbit, roughly 19,000 to 23,000 km altitude, is where GNSS constellations live and where O3b-class broadband closes its link budget. The radiation environment is punishing and the mission set narrow, but for those missions nothing else works.
Why navigation settled at 20,000 km and stayed there
The altitude band from roughly 19,000 to 23,000 km is not a compromise; it is the answer to a specific geometric problem. A GNSS receiver on the ground needs simultaneous line-of-sight to at least four satellites to compute a three-dimensional position fix. At low Earth orbit altitudes, each satellite is visible from any ground point for only a few minutes per pass, so closing that four-satellite window continuously would require hundreds of spacecraft. At geostationary altitude, the geometry degrades: satellites cluster near the equatorial plane as seen from the ground, the signal path through the ionosphere lengthens at high latitudes, and the propagation delay alone introduces roughly 240 milliseconds of round-trip latency that complicates timing-based ranging.
MEO threads the needle. GPS Block III satellites orbit at approximately 20,200 km in six orbital planes, giving a constellation of 24 satellites continuous global coverage with comfortable margin. Galileo sits at 23,222 km in three planes of eight. GLONASS uses 19,100 km. Each system arrived at a slightly different altitude through independent analysis of the same trade-offs: coverage geometry, atmospheric drag (negligible above 2,000 km), and the radiation environment discussed below. The convergence is not coincidence.
The Van Allen problem is not optional reading
MEO passes directly through the outer Van Allen radiation belt, peaking in electron flux around 15,000 to 20,000 km. The inner belt, centred near 3,000 km, is also a concern for satellites that must transit it during orbit raising. Total ionising dose for a satellite at GPS altitude over a ten-year mission can exceed 100 krad (Si) without adequate shielding, compared with perhaps 10 to 30 krad for a typical low Earth orbit mission at 500 km. Single-event upsets in unprotected logic circuits occur orders of magnitude more frequently than at LEO.
This has direct consequences for hardware selection and cost. MEO spacecraft routinely carry 5 to 10 mm of aluminium equivalent shielding on sensitive electronics, and their components are selected or screened to radiation-hardened or radiation-tolerant grades. Solar cells degrade faster: GPS Block II satellites used silicon cells that lost roughly 30 percent of beginning-of-life power over ten years; modern Block III uses advanced triple-junction cells with higher beginning-of-life efficiency partly to absorb that degradation budget. A government buyer considering a MEO mission must budget for rad-hard components, extended qualification testing, and heavier structures. These costs are real and non-negotiable.
O3b showed MEO could carry broadband, within limits
SES's O3b constellation, now part of the MEO layer of the O3b mPOWER system, demonstrated that MEO is viable for high-throughput communications. The original O3b satellites operated at approximately 8,062 km in an equatorial MEO orbit, using Ka-band spot beams to deliver round-trip latency in the 130 to 150 millisecond range. That is roughly half the latency of a geostationary link and low enough for voice-over-IP and interactive applications that GEO handles poorly.
The coverage geometry of an equatorial MEO orbit limits service to latitudes below approximately 62 degrees, which is a deliberate design choice for O3b's target markets rather than an oversight. A polar or inclined MEO constellation would cover higher latitudes but requires more planes and more satellites to maintain continuous coverage. O3b mPOWER, the follow-on system, uses larger satellites with digital beamforming to increase capacity, but the fundamental orbital mechanics are unchanged. For a sovereign operator, the lesson is that MEO communications is a viable architecture for equatorial and mid-latitude nations with high-throughput requirements, provided the satellite mass and radiation-hardening budget are accepted.
The honest limits: cost, complexity and a very short mission list
MEO is expensive in almost every dimension that matters to a national programme. Spacecraft are large. GPS Block III satellites mass approximately 3,880 kg at launch. Even the smaller O3b spacecraft came in above 700 kg. Launch costs to MEO are higher than to LEO because the delta-v requirement is substantially greater: reaching 20,200 km from a low parking orbit requires roughly 2.4 km/s more velocity than reaching 500 km LEO. Radiation-hardened components carry price premiums of one to two orders of magnitude over commercial-off-the-shelf equivalents for some part types.
More importantly, the mission set that genuinely requires MEO is small. Navigation at the system level requires it, but a sovereign nation building a GNSS constellation is committing to a programme of dozens of satellites, ground control infrastructure, and atomic-clock supply chains. That is a realistic ambition for perhaps five or six states globally. Communications at O3b-class performance requires it, but only if the latency advantage over GEO is operationally necessary and the coverage latitude suits the service area. Earth observation does not benefit from MEO: the altitude produces ground sample distances in the tens of kilometres for optical sensors, which is useful for meteorology but that mission is better served from GEO or purpose-built LEO constellations covered elsewhere in this library. A government that finds itself considering MEO for reasons other than navigation or low-latency broadband should examine those reasons carefully.
What a sovereign MEO programme actually requires
A credible MEO programme has four non-negotiable elements beyond the spacecraft themselves. First, a ground control network with high-elevation-angle visibility windows long enough to handle the slower apparent motion of MEO satellites; a single ground station is insufficient for continuous command and telemetry. Second, an atomic frequency standard supply chain, since GNSS requires on-board clocks stable to better than 10 nanoseconds per day, and only a handful of manufacturers produce space-qualified rubidium or caesium standards. Third, a launch procurement strategy that accounts for the higher energy requirement; not all launch vehicles can deliver meaningful payloads to MEO without an additional upper stage. Fourth, a radiation testing facility or access to one, since component qualification cannot be skipped.
Programme timescales are long. GPS took from the early 1970s to 1995 to reach full operational capability. Galileo's development from programme start to initial services spanned roughly fifteen years. A new entrant with access to modern components and existing design heritage could compress that timeline, but a realistic first-satellite delivery at MEO altitude, including radiation qualification, is unlikely to fall below four to six years from programme start. Sovereign control of the resulting system is achievable; the path is simply longer and more capital-intensive than for LEO missions.
Engineering parameters
| Altitude range (navigation) | 19,100 to 23,222 km (GLONASS to Galileo) |
| Altitude range (MEO comms, O3b class) | ~8,000 km equatorial |
| Orbital period | ~12 hours at GPS altitude; ~5 hours at O3b altitude |
| Total ionising dose (10-year mission) | 100 krad (Si) or above without shielding; drives rad-hard component selection |
| Typical spacecraft mass (navigation) | 700 kg to 4,000 kg depending on generation and capability |
| Round-trip signal latency (navigation ranging) | ~130 ms one-way propagation at GPS altitude |
| Round-trip latency (MEO comms) | 130 to 150 ms at O3b altitude; roughly half of GEO |
| Delta-v to reach MEO from LEO parking orbit | Approximately 2.4 km/s additional above LEO insertion |
| Ground sample distance (optical EO at 20,000 km) | Tens of kilometres; not useful for high-resolution observation |
| Minimum constellation size for continuous global GNSS coverage | 24 satellites (GPS baseline); practical systems carry spares |
One contract, one accountable engineer
Commissioned as one programme, not a stack of contracts: spacecraft, launch, ground segment, mission control, training and handover are priced together. Source-access terms and audit rights are agreed in writing before signature. Request a MEO mission feasibility review.