Satellite communications is the engineering of the space segment and the Earth stations that close the link: low Earth orbit (LEO) constellations, GEO and MEO transponders, ground station engineering at gateways and tracking dishes, inter-satellite links, and the phased array antennas that steer user and feeder beams. 3GPP now treats that space segment as a non-terrestrial network with transparent or regenerative payloads; ITU Resolution 35 times how NGSO filings must actually put satellites on orbit.
SIA counted 4,434 satellites launched in 2025 across 296 launches, a 65 percent jump on 2024, bringing the operational fleet to 14,266. The commercial satellite industry reached USD 303 billion inside a USD 429 billion space economy, satellite broadband subscribers passed 10 million, and the ground segment generated USD 165.2 billion . Demand is for people who have commissioned a live gateway or payload, not a constellation slide.
Challenges in Satellite Communications Recruiting
Low Earth orbit (LEO) constellations ITU Res 35 milestones are not a GEO fleet plan
Low Earth orbit (LEO) constellations turned launch cadence into a regulatory and operations problem. ITU Resolution 35 (WRC-23) requires non-GSO systems to file deployment information at milestones M0 through M3; for many assignments the second milestone fell due by 1 February 2026 . That is not a GEO frequency plan. GEO satellite transponders stay still. LEO and MEO vehicles move, so 3GPP's NTN overview lists moving cell patterns, higher Doppler and faster delay variation, and beam management that depends on ephemeris plus UE location .
SIA's 2025 launch count is the labour signal: 4,434 spacecraft in one year, with U.S. firms manufacturing 83 percent of commercially procured satellites . An engineer who has kept a GEO bent-pipe transponder healthy has not necessarily owned constellation station-keeping, intra-plane phasing, or a Res 35 deployment ledger. Treating "satellite" as one orbit class is how briefs fill with the wrong people.
Ground station engineering LEO gateways versus Estrack deep-space dishes
Ground station engineering is where the constellation hits dirt, and the dishes are not interchangeable. ESA's Estrack core is six stations in six countries, with 13 to 15 m near-Earth antennas and three 35 m deep-space antennas about 120 degrees apart in longitude. In a typical year the network provides more than 15,000 hours of tracking to 20 or more missions at above 99 percent service availability, on S- and X-band, with commercial partners (SSC, INTA, KSAT) filling LEOP gaps . CCSDS exists so those agencies can share tracking data; more than 1,000 missions have flown CCSDS-developed standards .
A LEO broadband gateway is a different machine: high-rate Ka or Ku feeders, dense pass schedules, optical or RF diversity, and often electronically steered apertures instead of a 35 m X-band dish. SIA's ground segment, USD 165.2 billion including GNSS equipment, is mostly user terminals and gateway electronics, not deep-space DSAs . An Estrack operator who has run a Cassini-class pass has not necessarily commissioned a LEO gateway under a constellation scheduler. A VSAT installer has not run CCSDS SLE cross-support.
Satellite transponders bent-pipe versus regenerative NTN payloads
Satellite transponders used to mean a bent-pipe: frequency convert, filter, amplify, repeat. 3GPP's NTN overview still names that non-regenerative payload, and notes that in transparent mode the NTN gateway at the end of the feeder link is the gNB, because the satellite is a remote radio head . Regenerative payloads add demodulation, decoding, switching and coding on board, so the satellite carries base-station functions and the gateway can be a router into the core . Rel-17 NTN was the first normative satellite access in 3GPP; later work adds L/S-band NR NTN bands and enhancements .
That split decides who can integrate D2D. 5G Americas describes Rel-17 NTN in sub-2 GHz for handheld devices and later Ka-band work for dish terminals at much higher rates . A bent-pipe payload engineer has not implemented on-board gNB functions. A 5GC engineer who has never closed a feeder link has not owned a transponder. The word "payload" on a CV does not record which of those two architectures flew.
Inter-satellite links require regenerative payload not a feeder-only hop
Inter-satellite links are the mesh that lets a constellation route around a missing gateway. 3GPP is blunt: usage of inter-satellite links requires regenerative payload . Optical ISLs are now a funded programme, not a paper. ESA's HydRON aims at a multi-orbit optical network with terabit-per-second class capacity; Element 1 is a LEO ring that relays with laser terminals, Element 2 extends to higher orbits, and ESTOL specifies bidirectional inter-satellite and ground-to-satellite optical links, updated to version 3.1 in June 2026 . HydRON Element 1's published baseline includes LEO nodes with multiple laser communication terminals at up to 100 Gbit/s for LEO-LEO connectivity .
An RF feeder-link engineer who has never aligned a laser communication terminal has not done ISL. An FSO terrestrial backhaul engineer (a different ITU-T G.641 problem) has not flown an LCT against a moving LEO neighbour. Probe PAT (pointing, acquisition, tracking), wavelength plan and whether the link was LEO-LEO, LEO-GEO or space-to-ground.
Phased array antennas user terminals versus gateway dishes
Phased array antennas are how LEO user terminals and many gateways keep a moving spacecraft in the beam without a mechanical dish slew. 3GPP's NTN overview already treats beam management as an ephemeris-and-location problem for LEO/MEO, with elliptic beam patterns on the payload side . 5G Americas separates handheld NTN (Rel-17 sub-2 GHz, shared-cell megabit-class downlink) from small-dish Ka-band terminals aiming at hundreds of Mbps . Those two antenna classes do not share calibration evidence.
SIA's 62 percent jump in satellite broadband subscribers, past 10 million, is mostly electronically steered user terminals, not 35 m Estrack dishes . A radar phased-array engineer can look adjacent on paper and still lack satcom scan-loss, grating-lobe and EIRP-mask work. A GEO VSAT installer who has peaked a dish has not calibrated a LEO user phased array through a pass. Ask which band, scan volume and whether the array sat on a user terminal, an aero/maritime terminal or a gateway.
Low Earth orbit (LEO) constellations operations evidence separates the NTN partnership announcement from the pass plan
Direct-to-device has become the headline that hides the operations job. GSA counted 123 publicly announced satellite-to-cellphone operator partnerships by end-June 2026, 23 of them with a launched service, inside 283 operator-satellite partnerships across more than 100 countries. Starlink led with 99 partnerships, then AST SpaceMobile (44), Amazon Leo (30), Eutelsat Group (29) and Lynk (21) . GSA's April 2026 market snapshot had already put 97 operators in 70 countries as investing in satellite-to-cellphone, with eight compatible chipsets .
A partnership announcement is not constellation operations. Rel-17 transparent NTN can ride a bent-pipe that a mobile operator never touches. Live LEO operations are ephemeris, gateway diversity, ISL routing, Res 35 milestone evidence and user-terminal phased arrays . An NTN protocol engineer who has never sat a pass plan is not a constellation NOC hire. A constellation flight-dynamics engineer who has never seen 3GPP SIB19 is not an NTN integration hire.
Gateway logs expose inflated ground station engineering claims
Satellite CVs inflate because the nouns are shared. "LEO" can mean a CubeSat lab, a single-gateway trial or ownership of a multi-site feeder network through a Res 35 milestone . "Ground station" can mean an Estrack 15 m LEOP dish , a Ka gateway, or a consumer phased-array terminal . "NTN" can mean a Rel-17 air-interface paper or a launched D2D service . "ISL" can mean a HydRON study or a flying LCT .
Useful verification is concrete. Which orbit and payload type, transparent or regenerative ? How many gateways, antennas or LCTs did they own? What EIRP, G/T, pass success or pointing error moved after their change? For phased array antennas, which scan volume and band? For inter-satellite links, was PAT closed against a moving neighbour ? For ground station engineering, was the interface CCSDS cross-support or a vendor gateway scheduler ?
The cost of skipping that probe is a dark feeder while 4,434 spacecraft a year are already competing for spectrum and sandboxes . A missed gateway commissioning slips D2D attach that GSA is already counting in the dozens of launched partnerships . False negatives cost too: a radar array engineer can look off-industry and still own the scan math a LEO terminal needs.
References
- Affordability and Productivity Drive Historic Satellite Industry Growth: 29th Annual State of the Satellite Industry Report — Satellite Industry Association (SIA). (accessed 2026-09-27)
- Non-Terrestrial Networks (NTN) — 3GPP. (accessed 2026-09-27)
- Resolution 35 (REV. WRC-23) — International Telecommunication Union (ITU-R). (accessed 2026-09-27)
- Estrack: ESA's global ground station network — European Space Agency (ESA). (accessed 2026-09-27)
- ESA begins next phase of fibre in the sky optical communications project with Canada — European Space Agency (ESA). (accessed 2026-09-27)
- ESTOL: ESA Specifications for Terabit/sec Optical Links — European Space Agency (ESA). (accessed 2026-09-27)
- 5G Non-Terrestrial Networks (NTN) — Global mobile Suppliers Association (GSA). (accessed 2026-09-27)
- GSA publishes new State of the Market report detailing global status of 5G networks, spectrum, devices and services — Global mobile Suppliers Association (GSA). (accessed 2026-09-27)
- About CCSDS — Consultative Committee for Space Data Systems (CCSDS). (accessed 2026-09-27)
- Update on 5G Non-Terrestrial Networks Briefing Paper — 5G Americas. (accessed 2026-09-27)
