Statistically insignificant warning?

[Lynn W Deffenbaugh (Mr)]

Greetings Hessu,

http://aprs.fi/info/?call=K6KUS shows the following warning message:

24 seconds between packets on average during 142 seconds.
This station is transmitting packets at a high rate, which can cause congestion in the APRS network.

I'm thinking that you might want to consider suppressing warnings if there's not a significant weight of packets behind them.  I mean, a 142 second sample hardly constitutes a risk to the APRS network!

Lynn (D) - KJ4ERJ - Author of APRSISCE for Windows Mobile and Win32

PS.  One of my users was concerned about this warning.  I really didn't go looking for something to report!

[doug.p...@comcast.net]

I was looking at the info on my beacon information and was presented with the following:

Packet rate:	18 seconds between packets on average during 919 seconds.

This station is transmitting packets at a high rate, which can cause congestion in the APRS network.

I agree with what the previous poster said about "consider suppressing warnings". APRS is not THAT busy and this shouldn't even be a concern. They want us to get involved and use APRS but then when you do they tell us we are using it too much! What the heck. It sounds like DMR, they want us to use DMR, but when we do use it, we then get lectured that we have to use a certain talk group, we can only use it for x amount of time, and we have to do this and do that. Its getting ridiculous... And 'they' wonder why the newly licensed hams (and hams in general) are not talking like they used to and the participation is becoming almost non existent!

Get rid of the warning messages...

-----------------------------------------------------------------------------------------

[Michael Jensen]

Hi 

Come on it is a warning. It tells you to consider if what you are doing are ok. Like any other warnings you can choose to ignore them. 

18 sec between packet are really a high rate. The RF channel can hold approx. 400 packet /10  min before high rate of collision happen. The assumption are that with TX delay and all, one packet take 2 sec for the ease of things. In many places one digi can hear 3 other. So when a user transmit one packet the RF Channel are busy for as long as 8 sec. So the first digi that received the packet is busy 40 % of the 18 sec transmit cycle allowing only for 60% for others. 

While 919 sec = 15.3 min are not a long period, the station have send 51 packets almost using 40% of the system capability in that time. If I saw that I would check if something in my setup have been corrupted. I will also read the full warning line. In the case Lynn are mentioning the monitoring time are 2 minuts.  A very smalle Sample period indeed. But on the other hand the warning is : Which might cause congestion in the APRS network. The reader are of the warning are to determine my them self. 

The messaged are very valid and every time it is shown, it should be at least consider. 

The impact of one packet to the network are at not always fully understod. Impact can be quite much higher then expected in areas with decent APRS coverage. 

73 de OZ1BZJ 

Michael 

--
You received this message because you are subscribed to the Google Groups "aprs.fi" group.
To unsubscribe from this group and stop receiving emails from it, send an email to aprsfi+un...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Bob Poortinga]

On Mon, May 21, 2018, 4:50 AM <doug.p...@comcast.net> wrote:

I was looking at the info on my beacon information and was presented with the following:

Packet rate:	18 seconds between packets on average during 919 seconds. This station is transmitting packets at a high rate, which can cause congestion in the APRS network.

I agree with what the previous poster said about "consider suppressing warnings". APRS is not THAT busy and this shouldn't even be a concern.

I consider this to be a rather uninformed statement. Has the writer ever run one of APRS-IS servers or a digipeater in a metro area? You may not hear much on your local RF frequency, but unless you have a high profile station or live in a very rural area, it is not possible to evaluate the "busyness" of the APRS network. The APRS-IS servers handle thousands of packets per minute and everyone has a duty to make sure that their station is not creating problems, either on RF or in the APRS-IS network.

The warning messages serve an important purpose. If anything should be done about this particular warning, it could be supressed if the number of packets is small or if the packets are old, for some value of "old." I would append "Increase the time between beacons to remove this warning" to the message.

Bob NG9M

[Doug Paquette]

I have my beacon intervals set for fifteen minutes when moving 2 mph or less, two minutes when moving between 2 mph and 60 mph, and one minute for speeds over 60 mph, and anytime I am making a turn of five degrees or more for five seconds it will beacon. I don't know where APRS.FI got the 18 seconds calculation but I feel my settings are acceptable, although APRS.FI doesn't think so.

In my opinion APRS in general is a poorly designed system. It should not be a simplex system and identification should not be determined by call signs and frequencies. There has to be a better way in regard to the functionality of APRS to solve the congestion/collision and duplicate id issues...

Have a great day.

73 = Best WisheS (not 73’s)

Doug

----------------------------------------------------------------

[Una Smith]

Doug,

If this beacon is transmitting VHF where several repeaters can hear it, it could be flooding the local VHF net to the point of severe obstruction of other traffic. That is because "smart" repeaters wait for other traffic to finish before transmitting, the hang time is a few seconds, and they repeat transmit some of the traffic they receive.

Is your beacon "smart" or dumb? Dumb means it transmits regardless of any other traffic and can interfere with other packets. Do you have control over the propagation rules on your packets?

Una KE5CEP

[Heikki Hannikainen]

On Mon, 21 May 2018, Doug Paquette wrote:

> I have my beacon intervals set for fifteen minutes when moving 2 mph or
> less, two minutes when moving between 2 mph and 60 mph, and one minute
> for speeds over 60 mph, and anytime I am making a turn of five degrees
> or more for five seconds it will beacon. I don't know where APRS.FI got
> the 18 seconds calculation but I feel my settings are acceptable,
> although APRS.FI doesn't think so.

Just to describe how it works:

aprs.fi looks at the latest raw packets from that station. For example,
these are Doug's packets:

https://en.aprs.fi/?c=raw&call=KD9IAQ-9&limit=25&view=normal

The rate warning on the info page counts in up to 50 packets, since the
latest packet. It'll stop counting when it sees a > 3600 second break
between two packets, and assumes the transmitter has been turned off; the
break over 3600 seconds should not be counted in the average packet
interval. It'll won't show the rate or any warnings if it finds less than
5 packets.

It's fairly easy to do the math from the raw packets by hand if you want
to double-check it:

* grab the timestamp of the latest packet
* go back up to 50 packets, but stop if you see a 1-hour (or longer) break
* count the number packets and the time difference between the first and last packet
* divide time difference between first and last packet by the number of
packets minus one (i.e. the number of gaps between packets)
* you'll get mean time between packets, which is shown on the info page

The raw packets display also shows the same thing for the currently
displayed packets. The above link *currently* says "25 seconds between
packets on average during 588 seconds", which will naturally be different
later on when more packets are sent.

Now, what exactly is "too often" will, of course, vary depending on the
amount of digipeaters and other APRS users in your area, and the
digipeater configurations. When you send an APRS packet and there are a
couple of digipeaters repating it, a single packet will take a few seconds
of channel time, and if everyone is transmitting every 20 seconds, there
can't be many APRS stations operating at the same time without a lot of
collisions (3, 4 or 5 cars maybe?). In many areas there are more stations
than that.

> In my opinion APRS in general is a poorly designed system. It should not
> be a simplex system and identification should not be determined by call
> signs and frequencies.

I agree, it could be much better; it was originally put together in a
rather hackish way with old hardware and firmware which was not designed
for this purpose. Unfortunately it's quite difficult to change it without
building a whole new parallel system. But, that's a whole new story; feel
free to raise that discussion on APRSSIG or elsewhere.

Michael, OZ1BZJ wrote:
> The impact of one packet to the network are at not always fully

> understood. Impact can be quite much higher then expected in areas with
> decent APRS coverage. 

I agree. Before discussing this topic a lot, it is a good exercise to
spend a few hours listening to the channel at a good, high digipeater
location, with a TNC / decoder monitor scrolling by, and looking at the
packets. There will be a lot of collisions heard, packets which are not
getting decoded due to those, and seeing and hearing the repeated packets
will help in understanding the effect of the transmit rate. You might not
get a good picture of the situation if you just listen to it in the car;
drive on top of the nearby hill first.

In some areas APRS is very busy. In rural areas it's not. A 20-second
average transmit rate may make it quite busy indeed, even with just a few
users in the same area.

Doug wrote:
> I agree with what the previous poster said about "consider suppressing
> warnings". APRS is not THAT busy and this shouldn't even be a concern.
> They want us to get involved and use APRS but then when you do they tell
> us we are using it too much! What the heck.

In some areas it is quite busy. The network has been, at times, in some
areas, been flooded by stations which do not have a polite configuration
which would leave channel time for other stations too. I think it's good
that at least somewhere you can get some feedback about the transmitter
configuration, in a consistent way, with some proof of what's actually
happened (the raw packets log).

It can of course be discussed what would be a good feedback text on the
site, and which would be the correct thresholds for the warnings. aprs.fi
currently gives a warning for an average distance less than 30 seconds I
think.

> It sounds like DMR, they want us to use DMR, but when we do use it, we
> then get lectured that we have to use a certain talk group, we can only
> use it for x amount of time, and we have to do this and do that.

Each network and protocol has specific properties, and I think it might be
good that users of these networks would learn from the feedback and
instructions. Talk groups are a great feature, as they allow people to
listen in to discussions of specific topics or geographical areas. They
should be used properly, as instructed by the people who take the time to
run the network for you. I think one should also be polite to the other
users of the network and not use the repeater all the time; let others
talk too - DMR repeaters only have two timeslots after all.

- Hessu