Details
A radio system today is a distributed application. Repeaters link over IP, dispatch positions are software, recording is a server, telemetry shares the infrastructure, and gateways join technologies that were never designed to meet.
What runs on the network
- Repeater linking — multi-site systems connecting sites over IP
- Dispatch positions — operator consoles with voice, position and incident logging
- Recording — time-stamped traffic retained for review and investigation
- Telemetry — SCADA and sensor data alongside voice
- Gateways — bridging DMR, TETRA and PoC into one operation
- Management — configuration, firmware and health monitoring of the estate
Each of these has requirements the office network was not built for.
What the network must deliver
| Requirement | Why |
|---|---|
| Bounded latency | Voice tolerates delay only up to a point |
| Low jitter | Variable delay produces audio artefacts that are hard to trace |
| Reserved bandwidth | Voice must not lose to a backup job |
| Stable timing | Multi-site needs it; simulcast needs microseconds |
| Availability matching the radio system | The link is now part of the coverage |
| Segmentation | Radio infrastructure on its own segment |
The one most often missed is jitter. A link with good average performance that occasionally stalls produces intermittent audio faults, and the investigation almost always starts on the radio side, where the cause is not.
Designing the two together
The common failure sequence: the network is built for office traffic, the radio system is designed afterwards, and the constraint is discovered at commissioning.
Designing them together means the radio requirement shapes the network — where links are needed, what availability they carry, what timing accuracy is required, which segments exist — while the network's real capability shapes what the radio design can assume.
Gateways between technologies
Mixed estates are the norm rather than the exception: DMR inside a plant, PoC for the fleet, TETRA for interoperability with public agencies. A gateway joins them so a talkgroup spans more than one technology.
What carries across is voice and basic group calls. Encryption, priority, telemetry and advanced data generally do not. A gateway is therefore specified as a voice bridge with its scope written down, rather than sold as a merge of two systems.
Failure behaviour
A well-configured multi-site system fragments rather than stops: each site continues serving its own area with reduced features while the link is down, and rejoins automatically when it returns.
That behaviour is a configuration outcome, not an automatic property. It is worth establishing before the day it matters — and it is one of the things a continuity review exists to check.
Security
Dispatch servers, repeater management interfaces and recording systems belong on their own segment, reachable by the people who need them and nothing else. They control the radio estate and hold recorded traffic, which makes them both sensitive and attractive. See firewall and network security.
TechnoRF designs, builds and maintains both layers as one system. Get in touch.
Frequently Asked Questions
What does the network have to guarantee for radio traffic?
Bounded latency and jitter rather than raw bandwidth — voice payloads are small but intolerant of variable delay. Multi-site linking also needs stable timing, and simulcast needs it to microseconds. A link that averages well but occasionally stalls will produce audio faults that are extremely hard to diagnose from the radio side.
Can radios from different systems talk to each other?
Through gateways, yes, and mixed estates are common: DMR inside a plant, PoC for the fleet, TETRA for interoperability with public agencies. What carries across is voice and basic group calls. Encryption, priority, telemetry and advanced data generally do not, so a gateway is designed as a voice bridge with its scope stated explicitly.
Where should dispatch and recording servers sit?
On their own network segment, reachable by the people who need them and by nothing else. They are the systems that hold recorded traffic and control the radio estate, which makes them both sensitive and attractive; leaving them on the general corporate network is the arrangement that looks convenient until it is examined.
What happens to a multi-site system if the link fails?
It should fragment into independently working sites rather than stopping. Each repeater continues serving its own coverage area with reduced features while the link is down, and rejoins automatically when it returns. Whether your system behaves this way depends on how it was configured — it is worth confirming before you need to know.