Latency in satellite systems: architectural selection criteria by use case
Propagation latency in a satellite link is a quantity determined fundamentally by the distance between the user terminal and the satellite, and cannot be reduced below the physical limit imposed by the speed of light. This constraint has direct consequences on the viability of certain applications depending on the type of orbit used. Systems in geostationary orbit GEO operate at distances of the order of 35,786 km, which implies round-trip delays of the order of several hundred milliseconds. Systems in medium orbit MEO present intermediate latencies, whilst constellations in low orbit LEO significantly reduce that delay by operating at altitudes of hundreds of kilometres. Stratospheric platforms HAPS, situated at around 20 km altitude, offer the latency values closest to those of conventional terrestrial networks within the non-terrestrial domain. For voice-over-IP applications and real-time interactive communications, the round-trip delay constitutes the critical quality-of-service parameter. Voice communication protocols are sensitive to accumulated delays exceeding certain perceptual thresholds, which limits the suitability of GEO systems for this type of traffic without specific compensation mechanisms. LEO and HAPS architectures are technically and functionally more appropriate for this use case, provided that geographical coverage and link availability are sufficient. In telecontrol applications and supervisory control applied to critical infrastructures, latency not only affects the operator experience but may also compromise the integrity of control loops. Industrial control systems requiring rapid update cycles, such as those associated with energy networks, water management or defence, impose latency requirements that must be evaluated according to the tolerance margin defined in the control system design. In these environments, the choice of orbital layer must be made following an analysis of the response time requirements of the control protocol in use. For bulk data transfer, such as Earth observation imagery, accumulated telemetry records or firmware updates to remote terminals, absolute latency has a lesser operational impact, as the dominant criterion is available bandwidth and the contact window. In this context, GEO systems offer structural advantages derived from their continuous coverage and aggregate throughput capacity. MEO systems present a balance between latency and coverage that makes them suitable for precision positioning applications such as GNSS and its differential correction services. The selection of architecture for a critical infrastructure operator or a public administration must begin with the functional specification of the service, including delay tolerance, availability requirements, geographical distribution of terminals and regulatory constraints on data sovereignty. No orbital or stratospheric layer is universally optimal: the engineering decision entails jointly assessing latency, coverage, resilience and user segment cost, and in many cases the most appropriate solution is a hybrid architecture combining capabilities from different layers.
NASSAT - Network Satellite Systems