Clock and Time Synchronization in a Private 5G Lab

[Pavel]

Short answer: Frequency accuracy keeps carrier and sample clocks within receiver tolerance; time synchronization aligns frames, TDD slots and distributed nodes. Use the level of synchronization required by the deployment—internal reference for a simple conducted cell, or shared 10 MHz/PPS/PTP/GNSS architecture for coordinated systems.

Before you start​

• Run private cellular experiments in an authorized, shielded or tightly conducted RF environment with sufficient attenuation between radio ports.
• Record RAN, core, UHD, firmware and operating-system versions together; a working configuration is a matched software and hardware set.

Step-by-step method​

1. Step 1: Identify whether the problem is absolute frequency, device time, frame alignment or network phase/time. Each needs different evidence and references.
2. Step 2: Verify the SDR reports reference lock and measure frequency offset against a known source. Configure the application to select the intended internal or external clock.
3. Step 3: For multiple radios, distribute references within input specifications and initialize time on the same PPS or supported network-time event.
4. Step 4: Observe cell synchronization, UE frequency correction, missed slots and long-run drift. Repeat after reboot because a one-time lock does not prove startup sequencing.

Concrete example​

A B210 can run a small single-cell test on its internal reference, but two coordinated cells may need shared frequency and time. Connecting only 10 MHz aligns rate but not necessarily device time or frame start.

How to judge the result​

Synchronization is adequate when the system remains inside frequency/timing limits under long-run load and all radios acquire the reference reliably at startup.

What to record​

• Troubleshoot registration in order: RF synchronization, broadcast system information, random access, RRC, authentication, security mode and PDU session.
• Monitor CPU scheduling, USB or network throughput, clock state and RAN logs during a sustained cell test.
• Use spectrum analysis and power calculation to confirm that leakage outside the controlled path remains below the lab requirement.

Common mistakes​

• Changing several hardware, software or RF variables at once, which removes the controlled comparison needed to identify the cause.
• Treating one autoscaled screenshot or one unusually good result as proof without recording the settings and repeating the test.
• Adding a GPSDO without configuring software to use it leaves the application on the internal clock despite the cable being connected.

Final check​

Cold-start the core, RAN, SDR and test UE in the documented order. A private-5G baseline is complete only when attach and traffic repeat inside the verified attenuated RF environment.