May I stick my nose in?So what if during forwarding the client(s) will be blocked. We just learn to live with it until we figure out how to overcome it.
DON'T THINK MESS. DON'T THINK MESS. DON'T THINK MESS. DON'T THINK MESS.Rather think of geographical area forwarding and/or most reliable station forwarding. Forward to the stations that you can contact the best (best signal and most of the time). Chose the best station, next best station, next best, etc. This was lessons learned by the Indian Military when they first started using communications. If you can "hit" a station 22-23 hours a day and have always strong signals and few repeats, that the station to forward to. In turn that station may have to re-forward the message to the "real" destination staion which may be only 100 km from you.
And always remember the "hidden transmitter". I will kill you on HF if try any kind of mesh network.
73, Walt/K5YFW Rein Couperus wrote:
Tnx for the comments Rich... well, this is the stuff I would really like to do. And the "working pause" out of the lab I have here will certainly help produce some interesting concepts. One problem is bandwidth/resources. During forwarding the client(s) will be blocked. Moreover, level 3 routing in a mesh produces quite some traffic in itself, if you are to prevent flooding. This quickly turns a mesh into a mess... But I am thinking about DTN (delay tolerant networking) in this respect, which is the way to go when we are really talking about ad hoc networking. This could be integrated with APRS stuff... knowing the geographic positions of your neighbour clients would enable intelligent geographic routing. I think all this could happen on a separate channel, so you could switch the client to 'mesh network mode' when you are not contacting a server. This could possibly free quite some resources... And we could use the rig control feature to create a 5-dimensional network, the 3rd dimension being frequency, and the 4th time schedules/modes... Stations forwarding could go to an assigned channel (up 500 Hz) and return to the service channel after they are ready. All this is only interesting if it is automatic. No fixed forwarding partners, no preset network topology map, with the exception of the working internet gateways. It will be a pleasure to do some conceptual thinking about it. We would need a lot of perl code, so if everybody starts buying the books... :-) 73, Rein EA/PA0R/P-----Ursprüngliche Nachricht----- Von: pskmail@xxxxxxxxxxxxx Gesendet: 23.01.07 01:51:46 An: pskmail@xxxxxxxxxxxxx Betreff: [pskmail] Re: BBS for emergency communicationsThis has raised a point that I have thought about some and would like to comment on. First during Hurricane Rita it would have been of first importance to be able to send either email or simple text messages between clients. This would allow some sort of message form to be followed. Second store and forward should be implemented between clients in the same sort of autonegotioate mode that mesh networking uses. That is if one of the clients on the air in the self forming net has a connection to an internet server then all mail is forwarded to that station. To do this first all stations on the common frequency would have to recognize each other and also who everyone can talk to as well as build a simple path [route] to each client. Something like that would be of great usefulness in an emergency. If each station builds a route list for the stations it hears and assigns a route index based on the amount of retries and broadcasts this list periodicaly then the rest would be straight forward. Rich Hudgins N5ale Ps I know its not as easy as I make it sound but it would be the first dynamic network on HF that I know about. -----Original Message----- From: pskmail-bounce@xxxxxxxxxxxxx [mailto:pskmail-bounce@xxxxxxxxxxxxx] On Behalf Of Rein Couperus Sent: Monday, January 22, 2007 2:06 PM To: pskmail@xxxxxxxxxxxxx Subject: [pskmail] BBS for emergency communicationsI have been looking into a BBS for adding to pskmail, to help those who* have no internet connection for the server * are forbidden to connect a radio to the internet (France) * are setting up emergency comms for cases when a large slice of the internet is down and there is no possibility to reach an HF gateway with running internet. Of course I have been shopping for something useable which is existing and alive. I have found only 2 cases of BBSes which are still in development/maintenance, and they are both german. The rest is either dead or DOS. (The last update of F6FBB BBS was in Jan, 2003). The best is probably the one from the Baycom group in Munich, called OpenBCM. But all these are cosmic solutions covering everything, the universe and the neighbours'cat. These BBSes are a nightmare to configure, and you need 2 people to run one. Moreover, the whole infrastructure these BBSesuse is dying quickly. In PA0 land the packet backbone has died, the local city node in Eindhoven has 1 packet link and 7 internettunnels. And the group of AX25 users has gone down from 85 to 4. The main concern I have is that it is all based on a fixed infrastructure, and I am sure what is needed in an emergency situation is ad hoc. The number of volunteers maintaining these systems is getting less in numbers, and the freaks running the infrastructure only talk amongst themselves.... Technical development has stopped. We are still looking at 9600 bd ax25 links. An on 30m you can hear the wonders of a full-fledged HF APRS network running with 300 Bd AFSK digipeaters!!! If we are then talking about voluntary resources, I think we need much more basic stuff which is very much automatic and can be set up in an easy but flexible way depending on circumstances. I had sworn I would not get involved in all this, but maybe it is worth getting started up. I have a nice test bed at the university of Eindhoven, who host a pskmail server AND a packet BBS (F6FBB). One of the first things we need to know is, what is the minimum functionality of such a BBS? To kick the ball off, I would say: * Storage of mail not deliverable via internet (immediate delivery in case of internet connection) * Store/forward of mail to a destination BBS via a HF link * User interface via Kmail or Evolution (POP3 interface) * Capability to send scheduled bulletins * Built in web server so people can also look at the mail locally (at the server location, which may be a command post of some sorts) via a connected LAN with a laptop This must be especially targeted at the emergency comms, as we have a satisfying working solution for normal situations. Ideas are welcome. Scenarios are welcome. Resources are even more welcome. 73, Rein PA0R -- http://pa0r.blogspirit.com -- No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.432 / Virus Database: 268.17.4/643 - Release Date: 1/21/2007 5:12 PM -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.432 / Virus Database: 268.17.5/645 - Release Date: 1/22/2007 4:10 PM