5G small cell synchronization - GPS vs 1588v2?

Deploying small cells indoors where GPS is weak. Is 1588v2 timing good enough for TDD small cells? What's the phase accuracy you're achieving?

For indoor TDD small cells, 1588v2 (PTP) can be good enough if the sync chain is designed for phase, not only frequency — but "GPS weak → just turn on PTP" often fails in the field.

What usually matters for TDD is absolute time / phase alignment across cells (often on the order of ~1 µs or tighter depending on the band and operator rules), not only ppb-level frequency holdover. That means: a grandmaster with a stable reference, aware boundary/transparent clocks where needed, and a backhaul that does not add huge PDV. On consumer or unmanaged Ethernet, PDV can blow the phase budget even when frequency looks locked.

Practical pattern: GPS/GNSS on the rooftop or a window-facing small cell as GM where possible; 1588v2 distribution indoors; holdover oven/TCXO sized for the outage you must survive. Measure time error with a calibrated probe or a known-good GM, not only "PTP locked" LEDs.

Open question: what time-error (ns or µs) are you measuring at the small-cell edge today on your backhaul, and is the network full on-path support or end-to-end over switches that are not PTP-aware?

— rfforum 编辑部 / Editorial Team