What is latency in satellite communications?
Latency is the time it takes for a signal to travel from source to destination. In satellite communications, the dominant component is propagation delay — the time for a radio wave, travelling at the speed of light (299,792 km/s), to traverse the distance between the ground and the satellite. A GEO satellite at 35,786 km adds 119 ms each way; round-trip time (RTT) is therefore 238–280 ms before accounting for processing delays.
Latency by orbit
GEO (35,786 km): ~240–280 ms RTT. Historically the dominant satellite broadband architecture; this latency limits interactive applications but is acceptable for broadcast, file download, and many enterprise applications.
MEO (8,000 km — O3b mPOWER): ~130 ms RTT — a significant improvement enabling VoIP and better interactive performance.
LEO (550 km — Starlink): 25–60 ms RTT — comparable to a fast terrestrial DSL connection, enabling cloud gaming, real-time video conferencing, and financial trading applications.
Additional latency components
Beyond propagation delay, satellite systems add: protocol processing at the gateway (1–5 ms), error correction encoding/decoding (variable), queuing delays in congested networks, and the RTT overhead of TCP handshakes. Satellite-specific TCP optimisation protocols (PEP — Performance Enhancing Proxies) mitigate the impact of high latency on TCP throughput, particularly for GEO links.
Why latency matters for the industry
LEO's low latency is the primary commercial differentiator over GEO for enterprise and consumer applications. It unlocks satellite backhaul for financial trading firms, enables satellite-powered cellular networks for operators running 5G NTN, and makes consumer satellite internet a genuine alternative to terrestrial broadband rather than a last-resort option.