Satellite Latency Comparison: GEO vs LEO vs MEO

Latency — the time delay between sending a signal and receiving a response — is one of the most critical performance parameters in satellite communications. The orbital altitude of a satellite is the primary determinant of signal propagation delay, and the three main orbit classes (GEO, MEO, and LEO) each produce fundamentally different latency characteristics.

Understanding latency across orbit types is essential for system designers selecting satellite architectures for latency-sensitive applications such as voice, video conferencing, real-time control systems, and interactive data services. This article provides a neutral engineering comparison of latency across GEO, MEO, and LEO satellite systems.

Glossary: GEO, LEO, Latency

What Causes Latency in Satellite Communications

Satellite communication latency arises from several contributing factors. The dominant component is propagation delay — the time required for an electromagnetic signal to travel at the speed of light between the user terminal, the satellite, the ground station, and back. Because radio waves travel at approximately 300,000 km/s, every additional kilometre of signal path adds measurable delay.

A typical satellite link involves a multi-hop signal path: the user terminal transmits to the satellite (uplink), the satellite relays the signal to a ground station or gateway (downlink), and the ground station routes the traffic to the terrestrial internet or destination network. The return path follows the same hops in reverse. The total round-trip time (RTT) is the sum of all propagation segments plus processing delays at each node.

Beyond propagation delay, additional latency sources include satellite transponder processing delay (typically 5 to 15 ms on bent-pipe transponders, potentially more on regenerative payloads), ground station processing and routing delay (5 to 20 ms), and terrestrial network transit time to the final destination. Atmospheric effects have negligible impact on propagation speed but can cause signal degradation that triggers retransmissions at higher protocol layers.

• Propagation delay: dominant factor, determined by signal path length at the speed of light
• Signal path: terminal to satellite to ground station to internet, and return
• Transponder processing delay: 5 to 15 ms for bent-pipe, higher for regenerative payloads
• Ground station and routing delay: 5 to 20 ms depending on architecture
• RTT formula: total propagation time (all hops) plus processing delays at each node

End-to-End Architecture | Ground Segment Reference | Glossary: Propagation Delay, RTT

GEO Satellite Latency

Geostationary Earth Orbit (GEO) satellites operate at an altitude of 35,786 km above the equator. At this altitude, the satellite's orbital period matches the Earth's rotation, causing it to appear stationary relative to the ground. This fixed position eliminates the need for satellite tracking and enables simple, fixed-pointing ground antennas.

The one-way propagation delay from a ground terminal to a GEO satellite and back to a ground station is approximately 120 ms under ideal geometry (directly beneath the satellite). In practice, slant-range geometry, non-equatorial terminal locations, and non-zero elevation angles increase this to 125 to 140 ms one-way. The full round-trip time — from terminal to satellite to ground station, through the terrestrial network, and back via the same satellite path — results in RTT values of 480 to 600 ms.

This high RTT is the fundamental tradeoff of GEO systems. The stationary orbit provides enormous coverage — a single GEO satellite can cover roughly one-third of the Earth's surface — and the ground infrastructure is simple and cost-effective. However, the half-second delay makes GEO unsuitable for latency-sensitive applications such as real-time voice calls (where delay above 150 ms one-way is perceptible) and interactive gaming.

GEO remains the dominant orbit for broadcast services (DTH television), wide-area VSAT networks, maritime and aviation connectivity, and backbone trunking for remote regions. In these applications, the coverage and infrastructure simplicity of GEO outweigh the latency penalty.

• Orbital altitude: 35,786 km (geostationary)
• One-way propagation delay: 120 to 140 ms (depending on geometry)
• Typical round-trip time: 480 to 600 ms including processing and terrestrial transit
• Coverage: approximately one-third of Earth per satellite; near-global with 3 satellites
• Ground infrastructure: fixed-pointing antennas, no tracking required

Maritime Connectivity Solutions | Energy Sector Solutions

MEO Satellite Latency

Medium Earth Orbit (MEO) satellites operate at altitudes between 8,000 and 20,000 km. This intermediate altitude produces significantly lower propagation delay than GEO while still providing broader coverage per satellite than LEO. MEO constellations typically use 8 to 20 satellites to achieve regional or global coverage.

The one-way propagation delay for MEO systems ranges from approximately 27 to 67 ms depending on the specific orbital altitude. Including ground station processing and terrestrial transit, typical MEO round-trip times fall in the 100 to 150 ms range. This represents a 3x to 5x improvement over GEO latency.

MEO satellites are not stationary — they orbit the Earth with periods of approximately 6 to 12 hours depending on altitude. This means ground terminals must track the satellite as it moves across the sky, and handovers between satellites occur as one satellite sets below the horizon and another rises. Tracking antennas or electronically steered arrays are required, adding complexity and cost to the ground segment.

MEO systems offer a practical middle ground between the coverage efficiency of GEO and the low latency of LEO. The moderate constellation size (compared to LEO's hundreds or thousands of satellites) keeps space segment costs manageable while delivering latency suitable for most interactive applications including VoIP and video conferencing.

• Orbital altitude: 8,000 to 20,000 km
• One-way propagation delay: 27 to 67 ms (altitude-dependent)
• Typical round-trip time: 100 to 150 ms including processing and terrestrial transit
• Orbital period: approximately 6 to 12 hours; satellites are not stationary
• Constellation size: typically 8 to 20 satellites for regional or global coverage

Glossary: GEO, MEO, LEO

LEO Satellite Latency

Low Earth Orbit (LEO) satellites operate at altitudes between 300 and 2,000 km. At these altitudes, the one-way propagation delay to the satellite is only 1 to 7 ms, resulting in round-trip times of 20 to 40 ms — comparable to many terrestrial fibre and wireless networks. This near-terrestrial latency is the defining advantage of LEO satellite systems.

Achieving global coverage from LEO requires large constellations — typically hundreds to thousands of satellites — because each LEO satellite covers a relatively small footprint on the Earth's surface. The satellites move rapidly relative to the ground, completing an orbit in approximately 90 to 120 minutes. This necessitates continuous satellite-to-satellite handovers (typically every 2 to 5 minutes) and sophisticated ground tracking.

Modern LEO constellations employ inter-satellite links (ISLs) using optical or RF crosslinks to route traffic between satellites without returning to the ground at each hop. ISLs can reduce total latency for long-distance routes because signals travel through the vacuum of space (where the speed of light is faster than in optical fibre) rather than through terrestrial fibre networks. For intercontinental paths, LEO ISL routing can achieve lower latency than the terrestrial internet.

The ground segment for LEO systems is more complex than for GEO. Terminals require phased-array or electronically steered antennas to track rapidly moving satellites, and the network must manage frequent handovers between satellites while maintaining session continuity. Gateway density must also be higher, as each satellite's coverage footprint is smaller and moves continuously.

• Orbital altitude: 300 to 2,000 km
• One-way propagation delay: 1 to 7 ms to the satellite
• Typical round-trip time: 20 to 40 ms including processing and terrestrial transit
• Orbital period: approximately 90 to 120 minutes; rapid satellite motion
• Constellation size: hundreds to thousands of satellites for global coverage
• Inter-satellite links (ISLs) can route traffic faster than terrestrial fibre for long distances

VSAT vs Starlink Comparison | Network Management Reference

Latency Comparison Table

Latency and Application Performance

The impact of satellite latency on application performance varies significantly by use case. Understanding these effects is essential for matching orbit selection to service requirements.

Voice over IP (VoIP) is highly sensitive to one-way delay. The ITU-T G.114 recommendation identifies 150 ms one-way delay as the threshold above which conversational quality degrades noticeably. GEO systems exceed this threshold by a wide margin (240 to 300 ms one-way), making real-time voice conversations uncomfortable. MEO systems (50 to 75 ms one-way) and LEO systems (10 to 20 ms one-way) both fall well within the acceptable range.

Video conferencing compounds the latency challenge because it involves both audio and video synchronisation, plus the visual feedback of seeing the other party react. Delays above 200 ms one-way create noticeable conversational awkwardness. GEO systems struggle with video conferencing, while MEO and LEO systems handle it well.

Web browsing performance is affected by latency through the TCP handshake and TLS negotiation process. Each page load typically requires multiple sequential round trips (DNS lookup, TCP SYN/ACK, TLS handshake, HTTP request/response). On a GEO link, these sequential round trips can add 2 to 4 seconds to page load time before any content begins transferring. LEO and MEO links experience this overhead to a much lesser degree.

SCADA and IoT control systems require predictable, low-latency communication for monitoring and command-and-control functions. While many SCADA systems are designed to tolerate GEO-level latency, real-time control applications (robotic teleoperation, autonomous vehicle control) require LEO or MEO-class latency. The bandwidth-delay product — the amount of data "in flight" on the link — also increases with latency, affecting TCP window sizing and throughput optimisation.

• VoIP: ITU-T G.114 recommends maximum 150 ms one-way delay; GEO exceeds this, MEO and LEO are within range
• Video conferencing: delays above 200 ms one-way degrade conversational quality
• Web browsing: multiple sequential RTTs for DNS, TCP, TLS, and HTTP add seconds on GEO links
• VPN and remote desktop: highly sensitive to RTT; GEO produces noticeable lag
• SCADA/IoT: tolerant of GEO latency for monitoring, but real-time control needs LEO or MEO
• TCP throughput: bandwidth-delay product increases with latency, requiring larger window sizes

How Satellite Internet Works | Satellite Link Budget Calculation

Latency and Network Architecture

Network architects employ various techniques to mitigate or work around satellite latency, particularly for GEO systems where the delay is most significant.

TCP acceleration (also called Performance Enhancing Proxies or PEPs) is a widely deployed technique on GEO satellite links. Standard TCP congestion control algorithms perform poorly over high-latency links because the slow feedback loop limits throughput ramp-up. PEPs intercept TCP connections at the satellite modem or hub, spoofing acknowledgements and using modified congestion control algorithms optimised for the satellite link characteristics. This can improve TCP throughput by 5x to 10x on GEO links.

Local caching and content pre-positioning reduce the effective latency experienced by end users. By caching frequently accessed content at the satellite terminal or hub, subsequent requests can be served locally without traversing the satellite link. CDN integration at the ground station level brings popular content closer to the satellite edge.

Local internet breakout allows satellite terminals to route internet-bound traffic directly to local internet exchange points or peering locations, rather than backhauling all traffic through a central hub. This reduces the terrestrial transit component of total latency and is particularly effective for MEO and LEO systems where the satellite propagation delay is already low.

Multi-orbit architectures combine satellites from different orbit types to optimise for both coverage and latency. Traffic requiring low latency (voice, interactive data) can be routed via LEO or MEO paths, while latency-tolerant bulk data and broadcast services use the broader coverage and simpler infrastructure of GEO satellites. Inter-satellite links between orbit layers enable seamless traffic steering based on application requirements.

• TCP acceleration (PEP): spoofs ACKs and uses satellite-optimised congestion control; 5x to 10x throughput improvement on GEO
• Caching and pre-positioning: serves frequently accessed content locally, avoiding satellite round trips
• CDN integration: positions popular content at the ground station or satellite edge
• Local breakout: routes internet traffic to the nearest exchange point, reducing terrestrial transit delay
• Multi-orbit architectures: steer latency-sensitive traffic via LEO/MEO, bulk traffic via GEO
• Inter-satellite links: enable traffic routing between orbit layers without returning to the ground

Network Management Reference | End-to-End Architecture

Summary

Orbital altitude is the primary determinant of satellite communication latency. GEO systems at 35,786 km produce round-trip times of 480 to 600 ms — adequate for broadcast, bulk data, and latency-tolerant VSAT services, but problematic for real-time interactive applications. MEO systems at 8,000 to 20,000 km deliver 100 to 150 ms RTT, suitable for most interactive applications including VoIP and video conferencing. LEO systems at 300 to 2,000 km achieve 20 to 40 ms RTT, comparable to terrestrial networks and suitable for all application types.

The latency advantage of lower orbits comes with increased infrastructure complexity. LEO systems require large constellations (hundreds to thousands of satellites), rapid handover management, and sophisticated tracking terminals. GEO systems achieve near-global coverage with just three satellites and simple fixed-pointing antennas. MEO systems offer a balanced middle ground.

No single orbit type is optimal for all use cases. Modern satellite network architectures increasingly combine multiple orbit types, using each where its latency and coverage characteristics best match the application requirements. Understanding the latency characteristics of each orbit class is fundamental to designing satellite communication systems that meet both performance and economic objectives.

Industry Solutions | How Satellite Internet Works