Part 4: How an NXDN Repeater Is Built

The only one of the six with both halves of roaming, and the beacon that makes it work. Working notes, v1.1, 10 September 2026.

NXDN is the only one of the six with both halves of roaming. It is also the mode where the amateur and commercial worlds are furthest apart, because the thing that makes the roaming work is absent from MMDVM entirely.

The same four questions are asked of every mode, in the same order, so the seven parts stay comparable:

The MMDVM route, and what gateway software it needs
The manufacturer route, with model numbers
Conventional or trunked
Roaming, and whether amateurs can use it

MMDVM route

MMDVMHost plus a board, gateway is NXDNGateway from G4KLX's NXDNClients. Needs an NXDN ID from the registry, separate from a DMR ID.

Manufacturer route

Kenwood NEXEDGE NXR-710 (VHF) and NXR-810 (UHF), later NXR-5700 and NXR-5800. Icom IDAS FR5000 and FR6000, later FR5300 and FR6300, plus the all in one CY5300 and CY6300.

The networking is a separate card, not standard. An amateur building one listed what it takes: "an ICOM IC-FR5000 (VHF), IC-FR6000 (UHF) with the IP Interface Card UC-FR5000 and CS-5000 Compact Flash Card," or "a Kenwood NXR repeater and IP Interface Card." One origin, older generation, but it is the first real answer to whether multi-site costs extra.

Do not assume every amateur NXDN repeater is MMDVM. It is an easy assumption to make from the hotspot world, and the forum post quoted above is an amateur setting out what building one on Icom FR hardware takes. That is one origin, and it is a build plan rather than a repeater confirmed on the air, so the honest statement is that the commercial route is being pursued by amateurs and not that anyone has counted what is running. The difference matters exactly here, because the beacon roaming below is available on that commercial hardware and not on MMDVM.

Conventional in amateur use

NXDN has trunking; amateurs do not use it. Kenwood sells it as single site, multi site and Enhanced Multi-site Trunking. It needs a pool of frequency pairs with one given over to a control channel that transmits continuously, the trunking gear is commercial only, and trunking exists to solve congestion that amateur repeaters do not have. The full argument is in part 7. No amateur NXDN trunked system was found anywhere.

Channel bandwidth

Mode	Channel bandwidth	Modulation	Bit rate
NXDN	6.25 kHz	4FSK	4800 bps
NXDN	12.5 kHz	4FSK	9600 bps

NXDN is not 4800 bps in both widths. It is 4800 at 6.25 kHz and 9600 at 12.5. Wikipedia states 4800 flat with no distinction and is wrong for the wide variant. The NXDN Forum's own overview, Ritron's datasheets and the emission designator 8K30F1E all give 9600 for 12.5 kHz. The standards body outweighs the encyclopedia.

NXDN chose to narrow the RF. At 6.25 kHz it genuinely occupies half a normal channel, so two NXDN channels fit inside one 12.5 kHz allocation. D-STAR made the same choice. DMR and P25 Phase 2 went the other way, occupying the full 12.5 kHz and dividing the time to carry two conversations. That is why NXDN has no timeslots to assign.

How NXDN assigns rooms, and the trap in it

Every mode has two different fields and they get confused constantly. An access field, which works like CTCSS and decides whether the repeater listens to you at all, and a routing field, which decides where your audio goes. NXDN is the awkward one, and it is the trap the whole distinction exists to catch.

Mode	Access field	Routing field	Range
NXDN	RAN	Destination ID	RAN 6 bit, dest ID 16 bit

RAN looks like a routing field and is not. It is access control, the exact analogue of a DMR color code, and MMDVMHost stamps its own configured RAN over whatever the radio sent before it reaches the network. That is why radio selectable modules need a reflector patch rather than a config change. The real routing field is the destination ID, which rides in bytes 7 and 8 of the NXDN packet and does survive.

Roaming POSSIBLE

Roaming needs two halves: a network half, several sites linked so that being on any of them puts you in the same conversation, and a radio half, something the radio can measure plus logic to move by itself. NXDN is the only one of the six with both. The verdict here is POSSIBLE rather than YES because both halves need commercial gear, which is the whole of the difference between NXDN and DMR in practice. The part 7 table uses the same word.

Repeaters transmit a beacon periodically on their downlink. The radio holds a list of those downlink frequencies, measures RSSI, and parks on the strongest. A hold timer stops it flapping between two sites of similar strength. The sites are linked over IP, which is why Kenwood files this under Conventional IP Network.

Kenwood: "As users roam through the network, subscriber units use beacon signals to select the best repeater." Icom, under LINQ Repeat+, calls it a vote scan: radios "listen to a set of downlink frequencies transmitted by the repeaters and automatically switch to the channel of the repeater with the strongest downlink signal," and "LINQ Repeat+ also includes a beaconing downlink signal sent by all repeaters periodically."

The beacon is not part of NXDN. The word does not appear anywhere in the NXDN Technical Specification, and its Table 2.2-1 marks Site Roaming "n/a" for conventional operation. Beacon roaming is a Kenwood and Icom system feature layered on top of the air interface. The clincher: Kenwood implements the same Site Roaming for DMR Tier II Conventional, and a mechanism that works across two unrelated air interfaces cannot belong to either one.

How the beacon works

The problem it solves. An idle repeater transmits nothing, so a scanning radio has nothing to measure. The beacon guarantees there is a signal on the air to measure.
What it is. A short transmission on the repeater's own downlink. Not a separate control channel and not a separate frequency.
It carries no site identity. Icom's own glossary is explicit: "Beaconing. A signal that is sent by a repeater that enables subscriber radios to receive a transmission for the purposes of subscriber voting." Something to measure, nothing more. The site list lives in the radio's codeplug, not on the air.
Periodic, not constant.
Output only. The repeater transmits it, the radio receives it. The input is never involved. The beacon path is entirely receive side: a radio roaming on beacons alone transmits nothing in order to roam. It transmits on the chosen site's input only when the operator keys up. That is true of the beacon, not of roaming in general. Kenwood also offers an active probe in which the radio does transmit, covered under Passive and active below.
The sites beacon together, coordinated off air. Kenwood nominates a Host Repeater that "manages the timing by which each repeater interconnected by using an IP network transmits a beacon," so that "multiple repeaters start transmitting a beacon almost simultaneously." The repeater front panel shows a Beacon Operating Status of Beacon Master, Member, Local or None. That synchronisation happens over the IP backhaul, which is another reason it cannot be an air interface feature.
Two sites need two pairs, and they must be different frequencies. The radio picks a site by measuring its downlink, so two sites on the same pair give it nothing to choose between. The same pair at two sites is simulcast, which is a different thing entirely.

Passive and active, and why every serious implementation has both

Roaming needs the radio to measure something. There are only two ways to guarantee there is something to measure.

Passive. The repeater transmits a beacon on a schedule and the radio listens and compares.
Active. The radio transmits a probe on the repeater input and listens for the repeater to answer.

Kenwood pairs its beacon with Active Site Hunt, where "the transceiver sends a Wakeup message to the repeater to prompt synchronization", governed by a Sync Wakeup Wait Time and a retry count. Motorola pairs Passive Site Search with Active Site Search. Anytone offers only the active probe, because a ham repeater historically does not beacon and its design cannot assume one.

An earlier draft of these notes said DMR and NXDN were exact inverses, one receive side and one transmit side. That was wrong on both counts and is corrected here. Kenwood's radio does transmit, and MMDVM's DMR repeater can beacon. The difference between the two modes is narrower and more specific than that: MMDVM implements the beacon for DMR and not for NXDN. It is an implementation gap, not a design philosophy.

The beacon interval, and why it is not printed here

The Kenwood parameter exists and is named Transmit Interval Time, set in the KPG-D7 programming software under Edit > Conventional IP Network > General > Site Roaming > Beacon. Its numeric range is not in any public document. It lives in the dealer-gated programming software help. Icom publishes no interval at all, only the word "periodically". Two field reports on forums give figures that do not cleanly agree, so no number is printed here. Getting it means a dealer with KPG-D7 open, or a capture.

What this means for an amateur NXDN system

Not reachable with MMDVM. Measured from the source: MMDVMHost's NXDN code has zero references to beacon, roam or site. Its message types are voice call, data call, status, remote control, auth and idle. Nothing broadcasts site information.

The Anytone AT-D890UV cannot roam on NXDN. Its NXDN Settings section contains only Roam Stay Time, 1 to 200 sec. No roam list, no site list, no vote scan. The real roaming controls, Roaming Zone and Auto Roam Start Cond., sit under Auto Repeater, which is the DMR side. A stay timer with nothing to roam to is not a roaming feature.

So the honest position is this. For a conventional amateur NXDN system of two or more repeaters, roaming is possible, and it works exactly as described above: each repeater beacons, and the radio picks the strongest. What it needs is commercial gear at both ends, Kenwood or Icom repeaters and matching subscriber radios. It is not reachable with MMDVM repeaters or with the amateur handhelds most people own. Possible, not practical for most.

On a multi-mode repeater

NXDN is single channel, so an NXDN-locked repeater carries one user. Its MMDVMHost TXHang default is 5 seconds, the same as P25 and a second longer than DMR and C4FM. The mode hang timers are in part 7.

Worth noting that an MMDVM multi-mode repeater carrying NXDN cannot beacon on it, so putting NXDN on a multi-mode machine forecloses the roaming described above no matter what the radios can do.

What is still open on NXDN

The numeric beacon interval. The parameter is named and located, Transmit Interval Time in KPG-D7, but its range is dealer-gated and not in any public document. Icom publishes nothing at all. Two forum figures do not cleanly agree. Getting this needs a dealer with the programming software open, or an off air capture.
The beacon's actual frame content. Established that it is not a defined NXDN message and that it carries no site identity. Not established is what the burst physically is: an idle message, a voice-shaped transmission, or preamble with nothing of interest. No source found decodes one. This also needs a capture.

Sourcing status for this page

NXDN conventional roaming: three origins, two of them manufacturer primary sources (Kenwood USA, Icom America, plus a RadioReference thread). Short of five.
The beacon being vendor design rather than part of NXDN: five origins, including the NXDN specification's own Table 2.2-1.
MMDVM sends no NXDN beacon: one source, but it is the codebase itself, which is the specification rather than an opinion. g4klx/MMDVMHost, NXDN sources, zero hits for beacon, roam or site.
Anytone AT-D890UV has no NXDN roam list: one source, the KD0PNQ programming guide, which is the specification for that radio.
NXDN multi-site needs a separate IP interface card: one origin, a forum post, older generation hardware. Weakest claim on this page.
The Kenwood beacon parameter name: one origin, the NXR-1700 Function Reference. It appears at three URLs but they are the same document, which is not three origins.
NXDN bandwidth: four origins per figure, including the emission designator and the NXDN Forum's own overview.