WA8LMF Mirror of WB4APR Website - 21 July 2008 FIXING LA

FIXING Los Angeles APRS

Automatic Position Reporting System

Forward: This page was written in 2005 and was part of the initiative of the New-N Paradigm. It is retained for historical purposes.

SUMMARY: Most of the APRS collision problems in LA can possibly be fixed by limiting ALL packets to ONE HOP in the greater LA basin! Simply take a look at this map and follow this line of reasoning.

ALOHA IS COMPLETELY QRM LIMITED: Notice that ALL of LA and its surrounding areas can theoretically be seen by one tall digi. Of course, due to shadowing, it does take a few more, but the reason we have such a small ALOHA user limit of typically 60 or so is because of the muliti-hops that every packet is taking! With just 2 hops, every packet is cluttering SoCal with potentially 16 copies of itself!!! because of all the digipeaters.

THE SINGLE DIGI CONCEPT: But applying the ALOHA calculation for a SINGLE digi in central LA, and one hop for everyone, the 144.39 channel could handle almost 180 users without problems! By making this digi full duplex, so that the channel is 100% available for uplinks, then a full 360 typical APRS stations could be accomodated with high reliabilty. This is more than everyone within 100 miles of central LA!!!

WHAT ABOUT ALL THOSE DIGIS?: The reason there are a dozen digis in the same area as can be handled by one, is because the channel is totally oversaturated by QRM by all the multi-hops and dupes and copies. Thus it is impossible to get into the high central site because of QRM. Thus, users had to put up more digis in their local areas so that their local signal was stronger than everything else the digi was hearing, and then they had to add another hop (quadrupeling the QRM) to get into the BIG digi! This process just snowballs to where the only way for success is more and more digis with more and more QRM and less and less range. IE, MELT-DOWN!

THE SOLUTION: Although we spent along time trying to solve the problems above with the SUPER-SITE concept and going to 9600 baud, it is immediately obvious that the SUPER-DIGI solution does not take advantage of the fantastic geography available in LA! That is, turning off ALL incoming packets from out of area and using a single BIG digi to cover the basin.

The problems to this simplistic KISS approach, however, are varried and real:

  • This would work for the LA basin only, but does not account for other hills and valleys and shadows
  • There is no easy way to get to this level of performance unless you can get EVEYONE to cut their path to 1 hop
  • There is no way to enforce this with the typical KPC-3+ or TRACE digi (Other than UIDIGI and DIGI_NED).
  • Its difcicult to get everyone marching to the same tune and listening to the same drummer at the same time

    THE REAL WORLD: Probably what would work well would be just a few digis that are optimally placed to cover the entire area. These BIG DIGIs would ONLY respond to ONE hop and ONE HOP ONLY. This would immediately eliminate all out of area QRM and also all ping-pongs in the basin and with a the ONE-HOP only system, success would return to APRS in LA!!! Here would be the phase in plan leading up to DIGI-DAY or "D" Day in LA:

    PHASE-1:

  • For now, choose a ONE-HOP name that is different from any existing APRS generic. I'd suggest NEWLA (for now)
  • Add NEWLA to the UIDIGI list of all digis (temporarily)
  • Add a Geographic ALIAS to all digis that best describes their coverage
  • SYSOPS decide what are the 3? MAIN digis which will cover the entire LA basin best.
  • Put BEACONS in ONLY those 3? digis (DIRECT) announcing this new 1-hop plan and a URL.
  • Make these beacons every 10 minutes DIRECT.
  • Also remove WIDEn-N entirely from these 3? special high digis (no one will notice since many others still do the job).
    At this point, the existing user has not been affected, but he is starting to be told that something big is about to happen. But since the big digis are sending these beacons every 10 minutes, DIRECT, this can give those skeptics a chance to re-learn what the real RF path of these high digis really are. They will be amazed. This will work, because the downlink from a DIRECT digi is orders of magnitude more successful in getting to a mobile on the ground with little QRM than the return path from the mobile to the high DIGI that lives in a world of 100% QRM.

    PHASE-2:

  • Now remove WIDEn-N and TRACEn-N from ALL digis within 100 miles of LA, and at the same time...
  • Add the UIDIGI Aliases of "WIDE2-2" and "WIDE3-3" so that users do not notice any difference
  • This enforces a 1 hop rule for almost everyone with no changes by anyone using typical paths
  • This also enforces a firewall all around LA protecting the big selected main LA digis from even hearing any out-of-area WIDEn-N packets
  • While it would be nice to use the UIDIGI ROM or DIGI_NED to enforce the 1 hop rule by eliminating any chance of using NEWLA, NEWLA, NEWLA etc... this is not required (see later)
  • Some digis may support the linear UITRACE LINKn-N part of New n-N paradigm for the I-5, I-10, I-15, I-40 and COASTn-N...
  • Although WIDE2-2 still works because it is an ALIAS focus users to use NEWLA to prove to themeslves that 1 hop works.

    PHASE-3, "D"-Day: At this point, everything in LA is operating at 1 hop using either "WIDE2-2" or NEWLA but there are still four copies of every packet flying around because every station not only hits his local digi, but also hits one or more of the big 3?. But this is much better than it used to be and so the BIG 3? are starting to hear mobiles further out because of less QRM. Now is the big day.

  • On "D" day, disable NEWLA in all low-tier digis in the entire area so that only the BIG 3? respond to it.
  • Suddenly the QRM in the whole area drops by a factor of 4 and the only remaining dupes are from the BIG 3?
  • Now from the low-tier digis, remove "WIDE2-2" and ALL GENEIRC APRS routing completely except for the geographic alias.
  • Now these local digis can serve as directed paths for people who need to send point-to-point messages but dont add general QRM
  • This system can support about 180 stations.

    PHASE-4: But this 180 stations is still about half of the potential full load within 100 miles of LA. To get to the ultimate 360, it is necessary to go to a full -duplex system so that the big 3? digipeaters themselves are not QRMing each other, but are providing a full 100% channel feed. To do this, each of the Big DIGIS should switch to a UHF transmitter and operate their TNC's in full duplex mode. Now each of the big 3 digis can EACH support a throughput of 360 users of every single station they can hear! Users still transmit on 144.39 but can tune in on UHF any of the BIG 3? they want to watch.

    WHAT ABOUT 144.39 OUTPUT? When the big digis stop half-duplex digipeating and only digipeat on UHF, doesnt this then prevent anyone running normal APRS on 144.39 from hearing all the activity?... YES and NO!!!

  • Fixed stations with home antennas will still hear many dozens of stations direct and will hear all their local ALOHA area direct becuase of the high density of APRS throughout LA
  • Mobiles who can only view 40 stations anyway, will now still be able to see everyone within local direct range which in LA can easily include 40 stations.
  • THIS is APRS at its best. Excellent LOCAL quality, Minimal QRM, yet full Internet connectivity.

    INTERNET and IGATES: Here is where some of the almost-abandoned low-tier digis make their best contribution. These digis also add an individual UHF full duplex AUDIO relay of their 144.39 RECEIVER. On this UHF channel, now, the local IGate has the full EARS on 144.39 for everone in range of that digi and can build the LOCAL LIST for that low-tier digi. This IGate then, with one-hop-only, and ONLY using the SPECIFIC digi callsign of that digi, can send any Internet traffic back to RF for any station in its LOCAL area with minimal QRM to the rest of LA. Since the IGate is using a directed path, there is just one and only 1 copy of this traffic. This is an ideal situation for IGates. Like the Cell phone system, they concentrate only on their small area and do not QRM

    LOW-POWER TRACKERS: Another application for the almost-abandoned low-tier digis is to QSY their input to the alternate APRS input channel for low-power trackers (144.99 if available in LA) These digis now only listen for low power trackers using a 2 hop path and digipeat them over onto 144.39. These digis should drop back to an output power of say, 5 watts because all they have to do is hit one of the BIG 3? on the second hop. The only alias they support is RELAY for the low-power trackers that are using the universal APRS path of RELAY,WIDE.

    WHAT ABOUT ABUSERS USING MULTI-HOP PATHS? Actually, 1 hop is ENFORCED in LA even thought the BIG 3? did not even have to upgrade to UIDIG ROMS! This is because, when they switched to CROSS-BAND full duplex digipeats, the BIG 3? no longer hear each other! Thus even an abusive path of NEWLA, NEWLA, NEWLA goes still only 1 hop via any one of the BIG 3?, but no farther.

    RESULT: APRS works as a LOCAL system throughout LA with maximum efficiency and can support about 300 stations without much QRM. APRS at its finest. Possible only because of the high density of population and the presence of very high digipeater sites. AND of course the use of ONE-HOP only FULL DUPLEX digipeating.

    THE FINAL PUNCH LINE: And saving the best for last, go back to step one, and instead of using NEWLA as the new special one-hop generic path for the LA BASIN, make it be RELAY in the first place!!!

    This is possible since RELAY has been banished completely in the past from LA so not many people are using it. Also then it makes LA 100% compatible with the rest of the world that is using RELAY,WIDE a the universal path. Also once you get out of range of the big LA digis, the next tier outward can be allowed to support RELAY and WIDE for up to 2 hops only. Also RELAY,WIDEn-N can even be supported with smart digis that truncate N to only N within N hops of LA or other city...

    The only down side will be any home stations that are still acting as RELAY and these will be impossible to track down since most of them do not do callsign substitution. But even this can be solved. Here is the plan:

  • First stick with the NEWLA plan to get everyone on the new page.
  • AFTER "D" day, send out search-and-destroy teams using the path of RELAY
  • As a fun exercise, they can track down, DF and use their smarts to find every remainig RELAY
  • In a friendly and gentlemenly manner, terminate these RELAY stations.

    Once all RELAYS have been exterminated, and not until then, add the RELAY ALIAS alongside the NEWLA in the BIG 3? DIGIS. Then you can keep NEWLA as the LA Basin home path, but then now LA fully supports the universal path of RELAY but at no expense to existing operations!!!

    Enjoy.
    de WB4APR, Bob

    You are visitor number: Since 6 Jan 2004..

    The Naval Academy is a registered user of APRS and WinAPRS. The purpose of this web page is to show several applications currently in use at this site and should not be considered as an advertisement or an endorsement of any commercial product. WA8LMF Mirror of WB4APR Website - 21 July 2008