RoIP (Radio over IP)

Two-Way Radio Concepts and Electronics Fundamentals · Last updated · About 6 min read

Definition

Radio over IP (RoIP) is the general term for carrying two-way radio voice and control traffic over an IP network (Ethernet, the internet, cellular, or satellite links) instead of over the air directly. Conceptually, RoIP is "VoIP with push-to-talk": it digitizes radio audio and PTT signaling into IP packets so radio traffic can travel over standard data infrastructure across long distances, the same way Voice over IP does for phone calls.

The key difference from ordinary VoIP is that RoIP must also carry the control signals that make a radio a radio: Push-To-Talk (PTT, the signal that keys the transmitter) and Carrier Operated Relay/Squelch (COR/COS, the signal indicating the radio is receiving traffic). A RoIP system that gets the audio right but the keying wrong is still broken.

Basic Architecture

A typical RoIP deployment has three parts:

The parts of a Radio over IP path and the signals it carries A donor control-station radio connects to a gateway, which digitises the audio and translates the push-to-talk and carrier-operated-relay signals; the gateway reaches dispatch clients and consoles across an IP network. Four kinds of traffic flow in both directions: voice packets, usually over UDP; PTT to key the transmitter; COR or COS to say the radio is receiving; and a control and status channel, usually over TCP, for channel changes, radio IDs, GPS and text. What a RoIP gateway actually carriesVoice is only half of it — the keying signals are what make it a radio and not a phoneDonor radioa real control-stationradio, on the airGatewaydigitises audio andtranslates PTT and CORIP networkLAN, WAN, internet,cellular or satelliteClient or consoledispatch software,desktop or mobile appRFOn the wire, in both directionsVoicethe audio itself, in small packets — usually UDP, because a late packet is uselessPTTkey the transmitter: the client presses, the gateway closes the radio's PTTCOR / COSthe radio is receiving: the gateway tells every client to unmuteControl and statuschannel change, volume, radio ID, GPS, text — usually a TCP sessionA system that gets the audio right and the keying wrong is still broken.Audio with no COR means nobody knows when to listen; audio with no PTT means nobody can answer.
This is the difference between RoIP and ordinary VoIP. A phone call needs only the audio; a radio channel needs to know who is keyed and who is receiving.
  1. A donor radio (control-station radio) — an ordinary mobile or base radio that provides the actual RF connection to the radio system or channel. It transmits and receives on the air exactly like any other subscriber radio.
  2. Remote clients and consoles — dispatch console software, desktop clients, mobile apps, or other gateways elsewhere on the network. When a remote user presses PTT, the gateway keys the donor radio; when the donor radio receives traffic, the gateway streams that audio to the clients.

Some deployments replace the donor radio with a direct wireline or IP interface into the radio system infrastructure (for example, a repeater's IP interface), but the donor-radio-plus-gateway model is the most universal because it works with any radio system a radio can subscribe to.

Radio-Side Interfacing

There are two broad ways a gateway connects to its radio:

  • Discrete audio and control lines (often loosely called a "4-wire-style" interface): separate wires carry transmit audio into the radio, receive audio out of the radio, a PTT line the gateway pulls to key the transmitter, and a COR line the radio asserts when receiving. This works on nearly any radio with an accessory connector, including decades-old analog equipment. Getting audio levels and signaling polarity right is the main installation effort — see Audio Levels and Radio Interfacing and COR and PTT Interfacing.
  • Digital control over USB or serial: modern radios (such as Motorola MOTOTRBO and Kenwood NEXEDGE models) expose a digital interface that carries audio and control together and adds features discrete wiring cannot: channel change, individual/private calls, caller ID (PTT ID), text messaging, and GPS data. The trade-off is vendor- and model-specific cabling and protocol support.

Detecting Receive Activity: COR vs VOX

The gateway must know when the radio is receiving so it can stream audio and show "channel busy" to clients. Two methods exist:

  • COR/COS signal: a hardware logic line from the radio that asserts the moment squelch opens. This is fast and reliable and is the preferred method.
  • VOX (voice-operated switch) / audio detection: the gateway watches the audio level and declares "receiving" when it exceeds a threshold. This is a fallback for radios that provide no COR line. It clips the first syllable, can false-trigger on noise, and hangs on after speech ends. If a RoIP system exhibits chopped word beginnings or phantom receive indications, check whether it is running on VOX detection.

Voice Transport, Codecs, and Bandwidth

UDP is used instead of TCP because live voice cannot wait for retransmissions; a late packet is as useless as a lost one.

The audio is compressed with a codec. Uncompressed-quality G.711 costs roughly 80–90 kbps per direction per channel once packet overhead is included; low-bit-rate codecs such as G.729 or Opus at voice settings bring that down to a few tens of kbps. Multiply per-channel bandwidth by the number of simultaneously monitored channels to size a WAN link. Codec choice also affects audio character — digital radio audio that has already passed through the radio's own vocoder degrades further with each additional lossy compression stage.

Latency Budget

Users judge RoIP by PTT-to-audio time: the delay from pressing PTT at a console until audio leaves the donor radio's transmitter (and the reverse path on receive). The budget is consumed by audio buffering in the gateway and client, codec framing, network transit, jitter buffering, and the radio system's own channel-access time (which on trunked digital systems can dominate everything else). Well-built LAN deployments add only tens of milliseconds; congested WANs, VPN detours, or oversized jitter buffers push total delay to where users talk over each other.

Unicast vs Multicast Distribution

When many clients monitor the same channel, the gateway can send each client its own copy of the audio stream (unicast) or send one copy to a multicast group address that the network replicates to all subscribed clients. Unicast works on any network, including across the internet, but bandwidth grows linearly with client count. Multicast is efficient on a managed LAN but requires network support (IGMP — Internet Group Management Protocol — snooping, multicast routing) and generally does not cross the public internet. Many products use unicast client connections with an optional multicast mode for large LAN dispatch floors.

Typical Deployment Topologies

  • LAN dispatch: gateways and consoles on one site network. Simplest and lowest latency; failures are usually cabling, IP addressing, or audio-level problems.
  • WAN / VPN remote sites: gateways at radio sites, consoles at a dispatch center, linked by private WAN or site-to-site VPN (Virtual Private Network). Watch latency, jitter, VPN MTU (Maximum Transmission Unit), and firewall rules.
  • Internet with port forwarding: a gateway behind a site router made reachable by forwarding specific ports to it. Workable but exposes services to the internet; a VPN is generally preferred.

Common Failure Classes

  • No connectivity: wrong IP settings, blocked ports, NAT (Network Address Translation) issues — work bottom-up per Troubleshooting RoIP and Gateway Connectivity.
  • One-way audio: firewall or NAT passing traffic in only one direction, or a broken audio pair on the radio cable.
  • Choppy or robotic audio: packet loss and jitter on the network path.
  • Distorted, weak, or humming audio: audio level or grounding problems at the radio interface.
  • Stuck PTT or phantom receive: wiring/polarity errors on PTT/COR lines, or VOX misbehavior.

Key Benefits and Interoperability Value

RoIP bridging devices/gateways can connect radios that otherwise couldn't talk to each other at all — legacy analog radios, modern digital (DMR/P25/NXDN) systems, and IP-based endpoints (like a browser client using WebRTC) — into one interoperable network. Reported benefits include lower infrastructure cost than dedicated wide-area RF links, improved reliability by leveraging existing data networks, and easier interoperability between otherwise-incompatible radio systems.

References

See also: CTI product documentation

This article is part of the CTI Products Knowledge Base. Questions about RadioPro or MCN products? Call +1 (513) 595-5900 or email info@ctiproducts.com.

Informational only. This article is general background and is not engineering, safety, legal or regulatory advice. Verify requirements with the equipment manufacturer and the applicable authority before relying on it. Product names and trademarks belong to their respective owners.