Neutral Host in a Stadium
- Martin James
- 10 hours ago
- 2 min read
“No roof access, no problem” timing for multi-operator 5G
Sector: Large venue neutral host (stadium)
Challenge: GNSS antenna installation blocked by venue approvals and limited roof routes
Requirement: Stable time/phase for multi-cell TDD operation (sub-microsecond-class alignment target)

Starting position
The neutral host operator had:
An in-building fibre network is already provisioned
Multiple radio zones with BBUs/O-DUs in secured comms rooms
Strict venue rules around rooftop penetrations and working at height
A rooftop GNSS antenna at each zone was not deployable in the scheduled window.
Timebeat deployment strategy
1) Centralise absolute time acquisition
A single accessible building riser allowed a short, approved antenna route to a headend comms space. A Timebeat “open time appliance” was deployed there as the primary timing reference (GNSS-fed).
2) Distribute PTP across the in-building transport
PTP was distributed over the venue network to baseband locations. The transport path was engineered to reduce worst-case PDV:
Timing VLAN separation
Queue prioritisation for PTP
Elimination of unnecessary contention points
3) Stabilise PTP behaviour in a mixed-traffic environment
Because event-day traffic spikes were expected, Timebeat filtering/servo behaviour was tuned for:
Rejection of delay outliers
Stable operation under transient congestion
Fast recovery after short disturbances
Result (operational outcomes)
Deployment schedule met without waiting for full roof approvals.
Baseband units synchronised reliably to the PTP source during both quiet periods and high-traffic events.
A practical operational model was introduced: alarms, monitoring, and acceptance tests focused on end-to-end timing stability, not just “PTP packets present.”
Why this matters to neutral host buyers
Venue deployments are timing’s worst environment: high utilisation, change control, and physical constraints.
The key lesson: you don’t need GNSS at every radio location. You need an architecture that acquires time where feasible and distributes it in a way that remains stable when the venue is full.