On Apr 2, 2010, at 11:00 AM, Kris Kirby wrote: > There is one issue that needs to be handled. When a D-Star repeater > "hears" another user on the input for a callsign not it's own, the modem > is captured, and any input packet that starts or is received in the > middle of that transmission is discarded. This, of course, would not be > permitted in the Motorola world; something would have to be done with > the received information, even if a band-opening allowed a remote > digital user to interfere with a digital trunking system's input > channel(s).
Not sure what you're talking about here Kris. (The squirrel dinner thing was very funny, BTW.) The D-STAR header ONLY... right at the beginning of the transmission, includes the four callsigns embedded in every transmission for source-routing purposes. The user enters them into the rig, and they get transmitted ONCE per transmission. Those callsigns are not INTERLACED throughout the transmission... So what REALLY happens is this... The D-STAR repeater detects a header that's not set to use the repeater's callsign, it's discarded. However, if a station can MANGLE their header, the last known good header is used. We see this on D-STAR on weak-signal users all the time in the logs. It's easy to "tailgate" in D-STAR. Additionally FM capture effect still stands, and if the bitstream isn't too mangled by the "noise" created by a "double", a station can talk over another. But it takes a VERY strong signal vs. a VERY weak one to pull it off. Additionally in the D-STAR specification, the header/routing information and the embedded 1200 bps simultaneous serial data stream that rides along with each voice transmission are NOT forward-error-corrected, while the VOICE portion (4800 bps) is. This causes all sorts of user confusion when they try out low-speed "background" data applications, or wonder why the callsigns showing up on the screen are all mangled from other stations, and/or their own transmissions won't route properly. "But everyone can HEAR me!"... yes, they can. But the repeater can't copy your weak signal well enough to dig out the callsigns in your header. Mix the tailgating and the latter, and you have confusion when weak stations participate in any activity that requires correct callsign routing from all participating stations... such as the almost-worthless Icom Multicast Routing for "linking" multiple repeaters. Obviously, almost all U.S. D-STAR repeaters also implement HARD linking via the add-on D-PLUS software, which works far better with weak user signals... but kinda defeats the whole purpose of having the callsign header/routing in the first place. So...... with all of that in mind, I don't know why you say the repeater will automatically discard interfering transmissions. It's far more likely that it will PASS them, if the header got mangled, as if it were a continuation of the first transmission... after the "double" stops occurring. > Practically speaking, I think that the earlier data should be thrown > out, and the packet decode restarted with the new signal. Of course, > short of doing SDR and de/re-coding on the fly, this is not a trivial > problem to fix. When your RSSI is measured as BER, it's a different > world. STREAM decoding is done continuously, on the fly on all of these systems. FM capture effect still happens. This is the design flaw that data engineers coming over to RF make all the time... and it shows the most in D-STAR. In P-25, the Unit ID information is continuously repeated (interlaced) in the frames, and if two radios key up at the same time... when the "double" stops, the decoder at the repeater can still figure out who's left transmitting. In D-STAR, the ASSUMPTION (and we all know what assumptions do...) was that the RF/Air interface would behave just like a hard-wired bit-pipe... and it doesn't. I don't have any knowledge of how MotoTRBO or the others handle this. Here's hoping they're interlacing the critical "must have" identification/authentication/routing information instead of treating it like it's not as important as the voice signal itself... p.s. Before anyone thinks this is a plug or a rant either way for or against any technology mentioned above, be advised that I've been working on bugs like this my entire career, and watching engineers make the same mistakes over and over and over again in the wireline telco world. I gave up picking favorites long ago... almost every data protocol out there sucks, in one way or another. I can explain how to break X.25, Ethernet, Gigabit Ethernet, Token Ring, IPX, IP, Q.931 messages... ahh, pretty much anything -- 'cause I've seen customers do it. Invent a better protocol, someone will invent a better idiot. (My favorite was two weeks ago when a worker for a VERY large telecom company wanted to know the "IP endpoints" of a hard down T1 voice circuit showing red-alarm at our equipment's end... meaning no PCM voltage arriving from the wires... at all.) I think we're trying to teach people this stuff so fast these days, they have no idea what you're talking about when you ask them to measure the voltage on a T1 circuit... or try asking someone to measure the voltage on Ethernet sometime. When it comes to RF digital protocols, the entire classroom full of hams would fall asleep long before you got past the basic framing of the "circuit", let alone talking about how a double at the RF/Air interface would affect it. Wireline techs have no concept that ELECTRICITY and all the E=IR and other properties that go with it... are actually traveling down those wires... they do seem to "get it" that when they replace it with plastic fiber optics it all works faster/better... but then they bend the cable beyond the bend radius allowed for the fiber, and wonder why it all falls apart again. NO CLUE about the physical world, just that you're supposed to plug it in and type some commands and it'll magically work for you. (Jeff KE9V did a GREAT podcast on this topic entitled, "If you don't know how it works, you can just ignore it" prior to him podfading and dropping the podcast, which really was excellent.) And they DEFINITELY don't understand the RF interference possibilities from such rapidly switched voltages. And (most) hams simply won't have the time in their lifespans to learn what needs to be known about digital RF systems. This is the dirty little secret of ham radio these days... we're not the old hams who knew the magic of RF and how to tame it, the majority of us won't make it above appliance operator stage. Thus, the folks who REALLY know it -- really need to try a little harder to make the on-air interfaces bulletproof, and ALSO to make sure the products are also released with TEST GEAR that any moron could operate. Seriously. We all know from reading here on RB that just understanding all the "gotchas" of FM analog, and repeaters, is many years of study. Add the requirement that the repeater operator also should magically understand routing protocols, IP, on-air framing formats, and all that jazz? It'll be a while. Ask any agency who deployed P25 when it first came out how many years it took their best techs to really UNDERSTAND what was going on in the system on a day-to-day basis... or if they even really believe they do, yet. I was going to say I wish I had a penny for every time I told someone to re-run their analog telephone lines away from the flourescent lighting ballasts in their drop-ceilings to get rid of their line noise, but... actually... they DID pay me for that. :-) Lack of knowledge of the basics, means job security for me... in the wireline world. They can continue to crank out dummies who can't spell "T1" for all I care at this point. It's dull teaching this stuff one-on-one during panic-stricken conference calls... but it's paid the bills now for quite a while. Starting to get off into the weeds here... sorry. Just lamenting how un-trained many MANY MANY "technicians" are these days. I'm sure their degrees and certifications will get them past HR departments easier than I, if and when I'm ever forced back out on the streets, hunting for a job... been there done that. The company brought in a trainer a few years back to teach "logical troubleshooting". Three clues in, I gave the answer. The trainer said, "How did you DO that?" Before I could reply, my boss (kindly) said, "We didn't hire him for his personality!" LOL! -- Nate Duehr, WY0X [email protected]
