This bridge connects M17 to the SvxReflector network — the linking system built into SVXLink, the long-running open-source repeater and node software. It's a close cousin of the M17-to-AllStar/EchoLink bridge, and it works on the same principle, but it reaches a different destination with a different personality: where AllStar and EchoLink link node to node, the SvxReflector network organizes everything around talkgroups.
SVXLink is software that turns a computer into a repeater controller or linking node. Its companion server, the SvxReflector, ties many SVXLink nodes together into one network. The M17-SVX Reflector bridge connects the M17 world to that network, so that an M17 user and the SVXLink/analog nodes on the reflector can share a conversation.
This is what gives the SvxReflector its own character, and it's the main thing that sets this bridge apart from the AllStar/EchoLink one. On the SvxReflector, traffic is organized by talkgroups — the same concept DMR users know. All audio sent to the reflector on a given talkgroup is retransmitted to every node that has selected that same talkgroup, and only one person talks at a time on each. Nodes can monitor a list of talkgroups and switch to one automatically when it becomes active.
The network also has a "QSY" feature: a station can call out on a broad talkgroup (a whole region, say), and once someone answers, move the conversation to a quieter talkgroup where only the interested nodes follow. It's a more structured, channelized way of linking than the connect-to-a-node model of AllStar or EchoLink.
The crossing happens, as usual, in the form of plain audio over USRP — the uncompressed audio-over-IP format the DVSwitch tools use. What's neat here is that SVXLink plays two roles at once, and that's what makes the bridge work without any extra glue.
Two logics, linked
An M17 stream is converted to plain USRP audio by a small program (USRP2M17). Inside SVXLink, one part (a "USRP logic") receives that audio, and another part (a "reflector logic") is connected to the SvxReflector network on a chosen talkgroup. Linking those two parts together inside SVXLink lets audio flow straight between the M17 bridge and the reflector talkgroup — in both directions. SVXLink is, in effect, both the bridge and the reflector's own native client at the same time.
So a transmission from an M17 user becomes plain audio, enters SVXLink through its USRP side, and is sent out onto the reflector's talkgroup — heard by every node monitoring that talkgroup. Audio coming back from the talkgroup makes the same trip in reverse and is re-encoded as M17.
As with the other M17 bridges, the conversion between M17 and plain audio is done entirely in software, because M17 uses the open-source Codec2 codec. There are no AMBE vocoder dongles to buy. A bridge from a proprietary mode like D-Star into the SvxReflector would need that hardware; an M17 bridge does not. M17's openness keeps this bridge light and inexpensive.
The M17-SVX Reflector bridge links M17 to SVXLink's talkgroup-based reflector network. It rides the same plain-audio-over-USRP path as the AllStar/EchoLink bridge, but its destination is a different kind of network — one organized into talkgroups rather than node connections — and SVXLink itself serves as both the bridge and the reflector client. Like every M17 bridge, it runs without special vocoding hardware, because M17's codec is open-source software.
This bridge rests on the work of several open-source authors. USRP2M17, which converts between M17 and plain USRP audio, is maintained in Tom Early's (N7TAE) nostar project. SVXLink and its companion SvxReflector network are the work of Tobias Blömberg (SM0SVX). The USRP linking that ties the two together depends on the UsrpLogic module contributed by Adi Bier (DL1HRC), which is not part of core SVXLink. This page explains how the pieces fit; their projects are where the software lives.
See: github.com/nostar/MMDVM_CM · github.com/sm0svx/svxlink · github.com/dl1hrc
A noncommercial hobby reference compiled by N6JET, gathered from public sources and shared freely for anyone interested in amateur digital voice.