# When GPS Goes Dark: Why Holdover Is Your Network's Insurance Policy
Every day, thousands of critical networks rely on GNSS signals — GPS, Galileo, BeiDou — as their primary timing source. It works beautifully. Until it doesn't.
And "doesn't" happens more often than most realize.
**The vulnerability is real.** GNSS signals arrive at Earth's surface at roughly -130 dBm — astonishingly weak. That makes them susceptible to:
🔹 **Jamming** — Cheap, widely available devices can drown out GNSS signals across kilometers. A single truck driver using a $30 jammer near a cell tower has knocked out timing for entire sectors.
🔹 **Spoofing** — Malicious actors broadcast counterfeit signals that mislead receivers into disciplining clocks to the wrong time. This is particularly insidious because the equipment may not even flag an alarm.
🔹 **Solar weather & outages** — Geomagnetic storms, satellite decommissioning, and even atmospheric anomalies can degrade or interrupt signals globally.
In 2022, European aviation experienced widespread GNSS interference near conflict zones. In 2016, a GPS satellite (SVN-23) broadcast an erroneous time offset for hours. Networks without robust holdover suffered cascading failures.
**This is where holdover becomes your lifeline.**
Holdover is the ability of a timing source to maintain accuracy *after* losing its reference signal. When a spec sheet says "holdover: <1.5 μs/24h," that tells you the clock will drift no more than 1.5 microseconds over 24 hours without GNSS. For 5G networks requiring ±1.5 μs phase alignment, that margin disappears fast.
**The holdover hierarchy matters:**
- **TCXO-based oscillators** — Drift of ~10-20 μs/day. Suitable for best-effort applications only.
- **OCXO-based** — ~1-5 μs/day. Adequate for many telecom applications with short outages.
- **Rubidium-based GNSSDO** — **<250 ns/24h.** This is the gold standard for mission-critical infrastructure.
A GNSS Disciplined Oscillator (GNSSDO) with a rubidium atomic clock doesn't just free-run during outages — it leverages long-term calibration data accumulated during normal operation to predict and compensate for drift. The result: sub-microsecond accuracy maintained for days, not hours.
**Real-world impact:**
A major European mobile operator deployed rubidium GNSSDOs across its 5G fronthaul network after experiencing 47 minutes of regional GPS interference. During a subsequent 36-hour outage, their rubidium holdover units maintained <300 ns accuracy — keeping all phase-synchronized services operational while competitors experienced handover failures and dropped calls.
A North American power utility substation, operating on legacy GPS timing alone, experienced a cascading protection relay misoperation during a jamming event. Post-incident, rubidium-backed holdover was mandated grid-wide.
**The takeaway:** If your network cannot tolerate even brief timing degradation, your timing architecture needs holdover — and the oscillator you choose determines your margin of safety. Rubidium GNSSDO transforms holdover from "hoping it's short" to "confident it doesn't matter."
Don't wait for the outage to discover your weakest link.
---
\#GNSSVulnerability #NetworkTiming #HoldoverCapability #CriticalInfrastructure #5GSync