Satellite RAN Integration

Also known as: Satellite Radio Access Network Integration, Satellite-RAN Integration

Definition

The architectural approaches for combining satellite connectivity with a radio access network, spanning satellite-based transport of RAN traffic between split points (fronthaul, midhaul, or backhaul over satellite) and direct satellite delivery of the radio access function itself under 3GPP NTN.

Key Points
  • Covers two distinct integration patterns: satellite-as-transport (carrying existing terrestrial RAN split traffic) and satellite-as-access (NTN, where the satellite is itself part of the radio access layer)
  • Satellite-as-transport is the more mature and widely deployed pattern, used for cell backhaul in remote sites for over a decade before 3GPP NTN standardization
  • Propagation delay budget is the primary design constraint distinguishing which RAN split points can tolerate a satellite hop
  • Requires careful timing synchronization since RAN procedures assume tightly bounded round-trip latency between split points
  • Delivers the most CE-relevant value in coverage extension and network densification for terrain or geography that rules out fiber or microwave backhaul
Concept

Where terrestrial fiber or microwave backhaul is unavailable or uneconomical, an operator can carry RAN traffic over a satellite link instead — most commonly backhaul, connecting a standard terrestrial base station to the core network via satellite rather than direct radio access integration. This has been standard practice for remote cell sites for years. The newer pattern, standardized as 5G NTN, makes the satellite itself part of the radio access layer, with UE attaching directly rather than through an intermediate terrestrial base station. Which pattern an operator chooses depends on whether existing terrestrial base station infrastructure is present to backhaul, or whether coverage must be delivered with no terrestrial radio equipment at all.

Explainer

The constraint that separates viable from non-viable satellite RAN integration is propagation delay tolerance at the chosen split point. Backhaul-over-satellite works because the RAN split occurs above the physical layer — the base station performs full baseband processing locally and only backhauls IP traffic to the core, which tolerates the roughly 500-600ms round-trip delay of a GEO hop reasonably well for most traffic types, though it still degrades real-time voice and any latency-sensitive control traffic. Fronthaul-over-satellite, by contrast, would require carrying raw or lightly processed radio samples between a remote radio unit and centralized baseband processing — a split with sub-millisecond latency budgets that no current satellite path, including LEO, can meet, which is why fronthaul-over-satellite is not a deployed pattern. Midhaul sits between these extremes and is viable over LEO for some functional splits but remains delay-sensitive. This delay-budget hierarchy is why direct satellite-as-access (NTN) exists as a separate standardized pattern rather than simply pushing the RAN split further toward the radio: NTN redesigns the RAN procedures themselves (extended timing advance, relaxed HARQ timing) to absorb the delay, rather than trying to fit a satellite hop inside timing budgets designed for terrestrial fiber. For CE verticals, mining and energy sites overwhelmingly use the mature backhaul-over-satellite pattern today because it lets them reuse standard terrestrial base station equipment, while maritime and aviation are the primary early adopters of direct satellite-as-access NTN because no terrestrial base station can be installed on a vessel or aircraft in the first place.