Telesat Lightspeed is a planned LEO broadband network using Ka-band ground links and optical links between satellites. This note focuses on Lightspeed and its development stages. It keeps the historical FCC design, later commercial plans and experimental NTN results separate, because their satellite counts and technical baselines differ.
- System Architecture
- Constellation Development and Deployment Stages
- Spectrum and RF Details
- Spectrum and Orbital Altitude Comparisons
- LEO 3 and 3GPP NTN Demonstration
- Technical Challenges and Design Tradeoffs
- FCC and Regulatory Baseline
- Reference
System Architecture
A satellite network needs more than a radio connection to a terminal. Traffic must also reach its destination through space or ground infrastructure. Lightspeed combines steerable radio beams, onboard processing and optical inter-satellite links for that purpose.
Terminal, space network and gateway
A dedicated terminal connects to a satellite over Ka-band. Traffic can then cross optical links between satellites before reaching a landing station. The landing station provides access to terrestrial networks and customer services.
Telesat describes digital beamforming arrays, regenerative processing and four optical inter-satellite links per satellite in its 2023 design overview. The optical links create routes between spacecraft instead of requiring every user connection to descend immediately through a nearby gateway. Telesat design overview, October 2023.
An optical route still needs a ground endpoint when the destination is on a terrestrial network. Routing also has to account for link availability and network load. More possible routes provide flexibility, but a route with more satellite hops is not automatically the lowest-delay path.
Broadband access and mobile-network backhaul
A satellite terminal can connect a remote site, ship, aircraft or terrestrial mobile base station. In a backhaul deployment, ordinary phones still connect to the local mobile network; the satellite carries traffic farther into the network.
This distinction matters when comparing Lightspeed with SpaceMobile. A Ka-band broadband terminal has a different RF front end and antenna from an ordinary cellular handset. The existence of a 5G waveform test over a satellite does not establish that every smartphone can communicate directly with the production constellation.
Constellation Development and Deployment Stages
Several different satellite counts appear in Telesat material. Each belongs to a particular filing, procurement plan or deployment stage. They should not be added together or interpreted as a current count in orbit.
Initial 2017 FCC configuration
The original U.S. market-access decision described a Canadian-authorized system of 117 satellites. It combined a near-polar group with a lower-inclination group. Both were LEO components, but their altitudes and plane arrangements differed.
Group | Planes and satellites | Altitude | Inclination |
|---|---|---|---|
Near-polar | 6 planes × 12 = 72 | About 1,000 km | 99.5 degrees |
Inclined | 5 planes × 9 = 45 | About 1,248 km | 37.4 degrees |
Total | 11 planes; 117 satellites | Two historical shells | 2017 baseline |
An orbital plane is the geometric plane containing the orbit. Satellites in one plane follow approximately the same orbital path, spaced along it. Inclination measures its tilt relative to the equator. Different planes can share one altitude. FCC 17-147, paragraph 2.
Later Lightspeed design and 2026 expansion
The later MDA-based programme differs from the 117-satellite FCC baseline. Telesat described 198 satellites in 2023, while the 2026 announcements discuss an initial deployment stage and a further funded expansion.
Dated source | Count or milestone | Interpretation |
|---|---|---|
October 2023 design overview | 198 satellites; approximately 1,300 km | Historical commercial design description |
March 2026 announcement | Initial 156 satellites with Mil-Ka capability | Initial deployment stage; not an in-service count |
August 2026 announcement | 69 additional satellites; initial network increased to 225 | 156 + 69 = 225 in that announcement |
August 2026 service forecast | Global service expected in Q1 2028 | Forward-looking schedule, not a completed service milestone |
Telesat identifies MDA Space as the builder for the expanded network. The August announcement links the expansion to an Arctic connectivity contract. It should not be used to claim that 225 satellites were already deployed. Telesat, 4 August 2026.
Spectrum and RF Details
The word Ka-band is too broad to identify an operating channel. The exact intervals, direction and authorization date matter. The table below starts with the explicit ranges in the 2017 FCC market-access decision.
Commercial Ka-band baseline
Earth-to-space means a ground station transmits toward a satellite. Space-to-Earth is the reverse direction. These regulatory ranges do not specify a single carrier bandwidth or divide every frequency between user terminals and gateways.
Direction | Ranges | Interpretation |
|---|---|---|
Space-to-Earth | 17.8–18.6 GHz; 18.8–19.3 GHz; 19.7–20.2 GHz | Three separate downlink intervals |
Earth-to-space | 27.5–29.1 GHz; 29.5–30.0 GHz | Two separate uplink intervals |
Satellite-to-satellite | Optical in the later Lightspeed design | Not part of the Ka-band RF table |
The gaps are deliberate: 18.6–18.8 GHz, 19.3–19.7 GHz and 29.1–29.5 GHz are not included in this baseline. Do not fill them simply because adjacent frequencies fall within the same broad band name. FCC 17-147, paragraphs 1–2.
Military Ka-band change in 2026
The March 2026 announcement changes the user-link frequency plan by assigning 500 MHz to military Ka-band. Telesat states that this replaces the same amount of commercial spectrum and leaves the gateway link unchanged.
This is a replacement within the announced user-link plan, not a statement that every link gains another 500 MHz. The reviewed announcement does not specify the exact replacement band edges. Therefore, this note does not infer those edges from a generic military Ka-band allocation. Telesat, 17 March 2026.
A spacecraft design announcement and a regulatory authorization answer different questions. The first describes intended hardware and service capabilities. The second defines permitted operations in a jurisdiction. A full current RF implementation also needs its applicable terminal, gateway and national permissions.
Spectrum and Orbital Altitude Comparisons
The comparison plots use dated baselines rather than invented Gen1, Gen2 and Gen3 labels. Spectrum colors identify direction. Altitude colors identify the historical near-polar shell, inclined shell and later approximate design height.
Color-coded frequency intervals
Each horizontal segment marks a frequency interval explicitly listed in the 2017 FCC decision. The gaps remain visible. The 2026 Mil-Ka announcement is discussed separately because its exact new edges were not provided.
Figure 1. Exact intervals from FCC 17-147, paragraph 2. This is a historical authorization baseline, not a complete current Mil-Ka frequency plan.
A colored interval is an authorization range, not the bandwidth occupied by one modem. The satellite can use multiple channels and beams within its applicable plan. Neither the length of a bar nor its color establishes a terminal data rate.
Altitude comparison and orbital planes
The first two markers represent the original FCC design at 1,000 and 1,248 km. The later marker shows the approximately 1,300 km height in Telesat’s 2023 design overview. These are different dated descriptions.
Figure 2. Dated baselines, not three simultaneously deployed generations. Sources: FCC 17-147 and Telesat October 2023 overview. No exact orbital assignment for the 2026 expansion is inferred.
The approximate later height is not an exact shell assignment for every satellite in the August 2026 expansion. The reviewed expansion announcement does not provide that orbital table. Similarly, the 117-satellite plane arrangement must not be copied into the 225-satellite programme without a supporting design source.
Higher altitude generally increases the area visible from one satellite for a given minimum elevation angle. It also increases propagation distance. This tradeoff helps explain why a constellation can use fewer satellites while still needing careful coverage, capacity and latency analysis.
LEO 3 and 3GPP NTN Demonstration
Telesat LEO 3 provides experimental evidence for 5G NTN over a moving satellite. The demonstration is useful for understanding timing and Doppler handling, but it should not be confused with the complete Lightspeed production network.
What the December 2024 test established
ESA reported a bidirectional Ka-band connection between ground-based gNB and NR NTN UE through LEO 3. The test used an Amarisoft protocol stack compliant with Release 17 NTN specifications.
The reported setup enabled 32 HARQ processes and adaptive modulation from QPSK through 64-QAM. Peak spectral efficiency reached 3 bit/s/Hz in both directions. The UE used SIB19 information, including satellite ephemeris, to maintain Doppler compensation and uplink timing. ESA, 23 December 2024.
What the test does not establish
The experiment used a defined ground setup and a satellite pass. It does not establish commercial handset compatibility, continuous global coverage or the final production air interface for every Lightspeed service.
Spectral efficiency and throughput are different measures. A value in bit/s/Hz requires an occupied bandwidth before it can be converted into a rate, and practical user throughput also depends on overhead and scheduling. No additional end-user rate is inferred here from the reported efficiency.
The presence of an NR NTN UE also matters. A test terminal with Ka-band RF equipment is not equivalent to an ordinary phone connecting without external hardware. Backhaul, dedicated satellite access and direct-to-handset access should remain separate when describing the result.
Technical Challenges and Design Tradeoffs
A broadband constellation must keep links usable while satellites, beams and traffic demand change. Lower delay than GEO is useful, but performance still depends on the entire route, terminal hardware and available radio resources.
Ka-band links and moving beams
Ka-band links need sufficient margin for atmospheric attenuation and rain fading. A moving satellite also changes the terminal look angle. Tracking, beam transitions and resource allocation must work together to maintain the connection.
A larger coverage footprint does not create unlimited capacity. Users within a service area share the available spectrum and processing resources. Concentrated demand can therefore create a capacity constraint even when several satellites are visible. Antenna gain, elevation angle and traffic loading belong in the same link and capacity analysis.
Optical routing and service continuity
Optical inter-satellite links offer additional routes, but the network must maintain those links and choose usable paths. Gateway availability, terrestrial interconnection and changes during satellite motion remain part of the end-to-end service.
A route can trade a nearby gateway for additional space hops. That may help reach a destination or avoid an unavailable ground link, but it can change delay and congestion. The useful metric is the resulting service performance, not simply the number of crosslinks.
Deployment is another constraint. Manufacturing completion, launch, orbital commissioning and commercial availability are separate milestones. A future service date should remain a dated forecast until actual service delivery is confirmed. The same distinction applies when comparing Lightspeed with already operating constellations.
FCC and Regulatory Baseline
The FCC document used here is the original 2017 market-access decision for a Canadian-authorized constellation. It provides a clear historical RF and orbital baseline. It does not by itself describe the entire later Lightspeed programme.
Market access and separate earth-station approvals
FCC 17-147 addresses access to the U.S. market under stated conditions. The decision also explains that corresponding earth-station applications need approval. Satellite market access is therefore not blanket permission for every ground installation.
Record | Identity | Use in this note |
|---|---|---|
Released 3 November 2017; S2976; SAT-PDR-20161115-00108 | Initial 117-satellite configuration and exact Ka-band intervals | |
2023 operator design overview | 198-satellite MDA-based programme | Architecture and approximate design altitude, not a replacement FCC grant |
2026 operator announcements | Mil-Ka change and 225-satellite initial expansion | Dated programme updates, not proof of completed deployment |
Interference coordination and source scope
The 2017 decision discusses ITU coordination, spectrum sharing and concerns raised by other operators. These topics explain why a frequency range cannot be interpreted as exclusive global access or an interference-free operating guarantee.
This note reviewed the introduction, background and coordination discussion through paragraph 8 of FCC 17-147. It also reviewed the relevant technical content of the linked Telesat announcements and the ESA NTN report. It is not a complete review of all later modifications, grant conditions or international operating rights.
The precise Mil-Ka band edges, a complete orbital table for the 2026 expanded programme and a final production modem configuration were not established by these sources. They are left unspecified rather than filled from generic band plans or older constellation parameters.
Reference
- FCC 17-147 - Telesat U.S. market access, 2017 (selected sections)
- Telesat - Lightspeed network architecture, 16 October 2023
- Telesat - Military Ka-band addition, 17 March 2026
- Telesat - Arctic contract and 225-satellite expansion, 4 August 2026
- ESA - Telesat and Amarisoft 5G NTN demonstration, 23 December 2024