"John D. Hays" <john@...>
I've been terribly busy of late, but thought I would take a few moments to share some thoughts on the topics in this thread.
It all depends on your application. The UDR56K-4 is a "platform radio" that can take on many different personalities and tasks.
RMS Gateway - critical mass 1, or 2 if you want to go above 1200 or 9600 bps. It can talk to existing RMS gateways and stations over the air and provide Internet connectivity to the rest of the Winlink/RMS ecosystem.
D-STAR Voice Half-Duplex Gateway - critical mass 1. The UDR56K-4 can support both a gateway and a RF module function in one package, attach power, antenna, and Internet. This will connect user radios into the whole D-STAR network.
D-STAR Data Access Point - critical mass 2. Since there isn't another DD unit that performs on 70cm and at the data rates as the UDR56K-4, at least 2 units are required for operation.
AMPRnet/Net-44 - critical mass 1. TCP/IP over AX.25 is available at 1200/9600 bps and as high as 56 kbps.
Mesh Networking/Self Healing Networks
There have been several people who have talked about putting HSMM into the UDR56K-4. I can report that olsrd compiles on the radio's ARM processor but we have not been able to do any testing at this point. I also don't have experience with the HSMM implementation to know if there are incompatibilities with standard OLSR. This is not a priority for initial release of the radio, however, it would be interesting to have someone in the community provide what differences there might be and possibly take up the port and support of HSMM to the radio. I have also taken a look at B.A.T.M.A.N. My main question about these protocols is can they scale back to to a 9600-56000 bps CDMA network? When a protocol assumes a high data rate transport, it just may be too "chatty" for a lower speed network.
Certainly some of these protocols could run on a WiFi dongle attached to the radio, or the radio could be part of the mix via Ethernet or other transport. I think a self healing network makes a lot of sense, we just have to keep in mind there are always tradeoffs. With the UDR56K-4 you are trading bandwidth and data rate for power and propagation advantages. This permits a network to be built with lower density of stations. The frequency of node beacons needs to be tempered by the speed and number of adjacent nodes -- so a mobile station can run further against a base station before it needs to switch base nodes, but conversely topology updates may be slower. It may be necessary for inventive hams to build up a light weight protocol that can create a balance between speed and network intelligence. Time will tell, but first we need to get the radio out with stated functionality.