Why my gateway receive only twice data

Good day,
I have a very big problem that I can’t solved, since two week and I will really appreciate your help.
I will try to be the more clear as possible. My gateway only receive the 2-4 first measures of the two nodes and at the same time, the measures are not transmitted any more (See more detail below).

Situation
I have two nodes, in an orchard, which measure the Soil Water Potential each hour. Between a measure, it goes to sleep because of ArdouniLowPower library.

My gateway was at 1 kilometer of the nodes and it now at about 50m from the nodes.

Here is the map
Yellow => are my two nodes
Violet => Actual location of my gateway
Blue => Initial location of the gateway
Pink => second location of the gateay
Red => Third location of the gateway
Green => Fothe location of the gateway.

Gateway
Until yesterday, I was using the following compenemt as a gatewa:

I have two same model. Into the orchard, the gateway is powered by a Gel 12V/100Ah with a 65W solar panel.

Yestderay evening I replaced that Gateway with a The Things Outdoor gateway, but that not fixed my problem.

The problem
During more than one year, the two nodes could tranmit the data to the first location of my gateway (Blue marker of the map). In spring 2019, I acivate the station/node and the iC880A-SPI gateway and the measure could be sent until the foliage appear in the tree. At the same time the gateway did not receive the data sent from the nodes. I was surprised that both nodes were not transmitting.
As the initial location of the gateway was not very good, I moved the gateway to the marker pink but I never got a reliable transmission. Most of the time, when the gateway start, only the 3-4 first measure was received, but once again, at the same time, the data was not transmitting. I moved again the gateway to the marker pink, red, violet, green and back to violet location.
Read location and green location of the gateway, the data was not translitted.

Then I took my home gateway and I replaced with the field gateway, and the problem was the same. With the home gateway or the field gateway, the gateway get only the 3-4 first transmission. I reinstalled 3-4 the gateway, I swapped the concentrator, the anetnnas and the Raspbery, but I always experiement the same problem.

I also tried with a third node, which transmittted each 30sec. The data were successfully transmit for about 1 hours.

I quickly suspect the foliage to be an obstacle because now the foliage are plenty but I really do not thinks, the foliage are the problems. This because, I move my third node in middle of the foliage where stand the second node, andthe data could be transmitted with a RSSI of -45. Morever

  1. One node are almost cover of foliage, a the second are not
  2. the second node is located at 20m of the gateway
  3. The data are sent with a RSSI of -45 and -85
  4. The data are not transmitted any more at the same time

I believe that if the foliage would be an obstacle, the nodes would not stop transmitting at the same time, but with a delay of days, particularely, because the second nodes has a bit of foliage and at 20m of the gateway.

I finally removed both gateway and I bought The Things Outdoor Gateway that I installed yesterday. I restarted one of the two node but I left running the first node. I started the TTOG gateway and only the two first measure was transmitted as you can see here. At the bottom you can see the RSSI.

I really can not understand why only 2-4 first measure are transmitted, first of all, one of the two node is the same that I used last year. The only difference, I update the program and added ArduinoLowPower library. Do you think that the ArdionoLowPower library make troubles?

Until yesterday, I suspected the two first gateway to be the issue, but now as I installed the new TTOG Gateway and I am still experimenting the problem, I have no idea where I have to look for the issue.

As I wrote, the first node is new, but the second is the same as last year.

Note: I am using ABP as activation method

Actually I absolutely do not know where I have to investigate and I wonder if the Arduino LowPower library make an issue, but I would be surprised because it’s the only difference between last year and this year, on node 2.

I would really appreciate if you can tell how you would investigate to solve this issue. This evening, I will go tho the field and inactivate ArduinoLowPower on one of the node, to see if it will work better or not.

Many thank and feel you free to ask me more detail if I missed some.

Old gateway in the actual location (violet) : http://www.smart-idea.io/temp/g/gateway.JPG
Old gateway: IMG_2478
Node 2 at 20m from the gateway IMG_2488
Node 1 at 160m from the gateway IMG_2432%202
TTOG gateway IMG_2568
Gateway location from Node 2 http://www.smart-idea.io/temp/g/gateway-from-st2.JPG

The thing very suprising is for me: Yesteday evening I restarted the second node while the first node was running since a couple of weeks, I started the TTOG. Node to transmitted a meaure to the gateway. Fortunately, the node 1 was “scheduled” to send a measure 5mn after, and he does. One hour later, both nodes transmitted a second measure and then, nothing else. Neither Node 1 or 2 transmitted a new measure to the gateway.

Then as I wrote, I do not know (and I do not think) the problem come from the node(s) and as now I have a new TTOG gateway, I do not think the problem come from the gateway, specially because the previous gateway experiemt the same issue. …

??

Apparently a measure has just been received to the gateway right now, from node 1 and 2

Honestly it seems a node issue.
Maybe related one of the things that could occur after long time, e.g., update of millis() during sleep or overflowing frame counter (did you check traffic on the TTN console and not only on the application? Look at the gateway traffic also. How many packets did you send with each node?).

Dear UdLoRa
Thanks. lot for your reply. That a good point, and I missed to write about. The gateway traffic is a lot and I was surprise. Where do come that traffic? The gateway is in middle of the winyard08

This morning, I also suspected the node, but as I wrote, the node 2 is the same as last year. I used RTC to schedule the measure. About the program, the main difference is ArduinoLowPower, because last year, the node was not program to sleep. Then I change a little bit to make the board sleeping when the measurement are not needed. But that’s a good point. It’s really the main difference. It’s look like, sInce I uploaded the new code with ArduinoLowPower, the problem came
.
It measure each hour, then sleep for about 55mn
How work my sleeping mode? It sleep for 8sec , wake up for a very short time, light a led, and sleep for 8 sec, wakeup for a very short time, light a led, and so on until next measurement schedule. So, in fact it never sleep more than 8sec (0r 10, I do not remeber)

Do you think that the sleeping mode make trouble?. It is the reason why I am going to unactivate it , later in the day and see how it work over the night

The application node traffic is null, but the gateway traffic is “little” huge. Any reason if no data are transmitted from my node to the gateway?

The gateway is in the middle of winyard, it should not have other nodes, than mine :blush:

Thank for helping

Gateways receive traffic that any node sends within a multiple kilometer range. Why do you think only traffic from your nodes would received?

The traffic page shows the gateway is not the problem. You should go back to last years code on your node to check if the problem disappears. Or get nodes with a certified LoRaWAN stack which will probably solve your issues.

I do not know which node you are using, however here you can find a possible reason for delays related to sleep: Adafruit Feather 32u4 LoRa - long transmission time after deep sleep

The other possible one is: do you have frame counter checks activated? after how many packets it stops?

Are you using OTAA or ABP on the nodes?

Dear All
Thank a lot for your reply and interresting answers.

Are you using OTAA or ABP on the nodes?

I am using ABP noe, but I am changing to OTAA

I do not know which node you are using

It’s an Adafruit Feather MO with LoRa but I added several compenment as RTC, EEPROM, etc…

do you have frame counter checks activated?

No I do not. On the TTN-Consol, I have check “Automatically update gateway”
But on the two application devises, the “Frame Counter Checks” are unchecked and “Frame Counter Width” is 32bit

Why do you think only traffic from your nodes would received?

That’s correct, I checked the Gateway map and there are two gatewaysat 3-4 kil from my gateway. So, I can guess there are nodes around my neighbour gateways.

You should go back to last years code on your node to check if the problem disappears.

I am agreed. The only difference is adding the ArduinoLowPower library and code. Yesterday evening, before leaving my job, there were a received message. Then I drove and when I arrived at the orchard to unactivate the ArduinoLowPower library (sleeping mode), there were four received message. Then I decided to not touch my nodes and leaving then running for the night. And until, the message are received (look at the graph loot at startion 1 and 2).
I wait until the messages are received before unactivating the Sleeping mode.

BUT, if you change the date to the 1 first of June (Look at the date field where is the word “de” and the word “Jusqu’à” (de => from, jusqu’à => to), you will see that aleay the nodes transmit or not at the same time, like if there were synchronized!! (In my graph, if you look at the second station, you have to set the date again. I will change it later)

For exemple, yesterday, at 12:53 UTC, the gateway received message from the two station at the same time. How is possible that the transmission of the two node are “synchronized”, while the first node is running sing 2 weeks, and I restarted the second node. Both node have their own RTC, Radio, Battery, solar panel, etc. The two station are aurnom. Is the reason why I ma still focus on the gateway even if I understand your comment.

When the TTOG gateway will not received message anymore, I will unactivate the Sleeping mode on one of the node and see how work the two nodes.

I also will investigate on the link that UdLoRa provided and sleeping mode suggestion, but if you have any clarification about what would be reason that the transmission of the two node, stop and start always at the same moment.

Thanks,Cheers
Pierrot

If vineyards are on hills, you are in a privileged position for receiving a lot of traffic from the flat land :slight_smile:
I am not far but also not too close to vineyards and my packets are usually received by gateways as far as 30km.

In fact, this land “is not really on hills” is trange to say like this :grinning: My area is fortunately more or less flat . It’s not like other area of my country like here where is really sloping :grinning:

If vineyards are on hills, you are in a privileged position for receiving a lot of traffic from the flat land

Yes, but in my configuration, even if my area is not very sloping, the gateway is just above tree trees, but do not dominate the horizon (I do not know if I can wrote it like this in english)

30km is really a lot, that great :grinning:. If onece I can move my gateway to the highest level of my area, it would be the best :grin:

When the farest station send a message, the RSSI is -80dbi. Is -80dbi acceptable? from which dbi, the message could not be sent any more?
If you gateway receive message from your nodes from about 30km, which is the DBI? Did you check it?

It depends on SF; at SF12, the least I saw was RSSI -122, SNR -19. -80 is sort of powerful signal, the node is not too far.
Consider that in the gateway you see not only LoRaWAN traffic, but also generic LoRa and even 868MHz remote controls. However, when there is a join request, likely it’s LoRaWAN. However, likely they do not rely on your gateway only, so check on the map or on TTNMapper for other gateways around (although they could be on other networks, so you cannot see all of them). On the map there are 12 gateways around Geneva…

Anyway, all of this is not related with your issue, that likely has to do with the node.

Dear All,
I come back with this discussion because my problem come back but there are some new.
But let resume shortly,

I have a now a great gateway The things Gateway Outdoor, and it has been working well for 1 month.
I do have 4 nodes. 3 of there are in the orchard, at about 500m from the gateway. The fourth is at 1m of the gateway. The fourth is used the monitor the Gateway battery level. Each nodes transmitted about 45bit of message (Payload), each hour.

The 4 nodes was transmitting very well the message to the gateway for one month, until yestersay, Yesterday at 12.00 (UTC) all of the 4 nodes stop transmitting.

Yesterday evening, I went to the Gateway, I powered off the node 4 and powered it on. (The node 4 is at less 1m from the Gateway). But now way, no message are transmitting after I truned Node 4 on.
I powered off the gateway and powered it on. I reseted the “frame counters” of node 4. I reseted aganin the node 4. But now message are transmitted.

How can it possible? Node 4 worked very well for one mothn since I powered.

And Why the four notes stop transmitting at the same time? (there are not been tunred on at the same time)

Can it be possible that I exceed some limitation? How can I check it?
It is incedible :slight_smile:

Note: At TTN console, there is traffic on my Gateway. I do not where does it come, but there not from my nodes

Thank for helping

  1. Please do not post the same message multiple times.

  2. Why do you suspect your gateway? As it is receiving data from other nodes it seems to be working alright.

  3. What causes you to state 4 nodes stop transmitting? Is it because you do not receive data in your application?

Not getting data in an application can be caused by issues in the back-end. Have you ruled out possible issues in the back-end?

no… not from what you stated ’ 4 nodes- transmit about 40bit of message (Payload), each hour ’

Hello! thank for your reply.

I am really confused with that issue! Before June, I had two gateways, I “built” my self and I started experiencing that issue. From middle of June, I replaced it with the TTOG gateway and since it worked like a charm until yesterday… and I experienced the same issue :frowning: I do not kow where to investigate.

You are right, the gateway works fine. As you said, there are traffic from other nodes than mine.

I just returned from the Orachard. In the Orchard, there are 3 nodes I built. Each have a LCD display and the Tx are display. The node 1 showed about 730 Tx. The node 2 and 3 (there are beside) showed about 250 Tx. It’s clearly means, Node 1 started several days before nodes 2 and 3.

As I wrote, my TTOG gateway is about 500m. Some weeks ago, I added a 4th node beside to the gateway, only to monitor the battery of the gateway. I have not built it. It’s a Adafruit Featherwing with a Adafruit Feather MO Basic. That 4th node worked immediately and as it located less than 1m to the gateway antenna, I see no reason, the gateway can not see it. No reasons that the messages from node 4 are not transmitted to the gateway (and then save to my MySQL database)

The case which really trouble me (once again) why the 4 nodes are not seen from the gateway from the same time??? I means, each nodes (the 4) take a measure each hour and send it to the gateway. If you look at the charts , at the dashboard, you can see the “Dernière mesures” (Last measure). All of the 4 nodes time is between 11:50 and 12:39 (UTC). You also can click on Détail to see the charts (and the time) of each points.
You can also see, from 12:39, the messages of all of the nodes could not be received.
No payload (or messages) has been received from the Gateway and sent to TTN … and not saved to my MySQL DB. Do see what I means?

That’s really trouble me, why at the same time (within 1 hour) the messages of the 4 nodes (even the node 4 which is beside of the gateway) can not be transmitted. And it’s the reason why, I do not suspect and software issue from my nodes. Otherwise, the data would not be unsent within the same hour, after 1 month of good activity. Do you see what I means, in spite of my french-english :slight_smile: :slight_smile:

I will return to my orchard, tomorrow to check if the Tx increase, on the LCD display , but I am confident.

Please do not post the same message multiple times
That’s the reason why I suspect a network issue and I created a new post in Network & routing category after deleting my today’s post (which was not deleted). But it was not a good idea from me, sorry :slight_smile:

What causes you to state 4 nodes stop transmitting? Is it because you do not receive data in your application?

Yes.
From my TTN console, I go to my Application where are registered my 4 nodes. The 4 nodes are not seen since yesterday at around 12:00 (UTC). The charts stop displaying result, as you saw. And of course, no data are transmitted to my server/MySQL. I would think differently if at least, two node were transmitting.

Then could it be a issue from the TTN console?
Something can interfere with the good data transmission between my nodes and my bridge, despite the fact that I am in the nature, in the middle of the vineyards?

I am really confused. I am confident with my nodes, I am confident with my gateway…

I hope, I gave you good information and my english was not too bad
Thank for sharing your point of view

There have apparently been some issues lately with the TTN infrastructure, and you might simple be seeing that

This can be problematic: a node so close likely blanks out the ability of the gateway to receive from comparatively more distant nodes even on other channels. If you are going to do this, you want to keep the duty cycle of that test node very low, as your gateway is effectively disabled for other nodes any time it transmits.

Hello, thank for your reply. It a very interesting what you wrote. I am not at home right now but I will be back with more detail, I have not thought to write on my previous post. Thank a lot!!!

That is too close to the gateway. The signal will be too strong resulting in frequent CRC errors which result in packets being discarded in stead of forwarded to your application.

This can be problematic: a node so close likely blanks out the ability of the gateway to receive from comparatively more distant nodes even on other channels.

Ok, that is interesting, but do you think, it’s the reason of issue? If yes, why it works for a couple of weeks and why I experienced the same issue in May-June while no node where so close to the gateway?

My gateway it on my trailer
IMG_2669
The node 4 is inside of the trailer, close to the door and connected to the 12V/90Ah battery through a PoE injector/adaptator 12V->48V. I will disconnect it tomorrow.
Then what should be the minimum distance of a node to the gateway antenna (I could move it to the other side of the trailer.

If you are going to do this, you want to keep the duty cycle of that test node very low, as your gateway is effectively disabled for other nodes any time it transmits.

Should I make the node4 measuring and sending data, 30mn after the node 1,2 and 3?

There have apparently been some issues lately with the TTN infrastructure,

There is way to check that? This can be interesting in the case, next time, if it’s happen again, I could know immediately.

Other information:
¨At TTN console, I created an application and registered 4 nodes. Into the integration tab, I added HTTP integration with the following setting

and here is my paylod format

function Decoder(bytes, port) {
  var str=String.fromCharCode.apply(null,bytes);
  var astr = str.split(",");
  
  var regex = /([a-z]+)(\d+)/g;
  
  //var resultats = [];
	var paire;
 
  var decoded = {}
  var te,pr,hu,lu,w1,w2,w3,da,ts,ga,ba,ar,an,it,wr;
  while (null !== (paire = regex.exec(astr[0]))) {
		//console.log(paire[1]);
		switch(paire[1]){
		  case 'a':
		    //console.log("te",paire[2]);
		    te=paire[2];
		    break;
		  case 'b':
		    //console.log("pr",paire[2]);
		    pr=paire[2];
		    break;
		  case 'c':
		    //console.log("hu",paire[2]);
		    hu=paire[2];
		    break;
		  case 'd':
		    //console.log("lu",paire[2]);
		    lu=paire[2];
		    break;
		  case 'e':
		    //console.log("w1",paire[2]);
		    w1=paire[2];
		    break;
		  case 'f':
		    //console.log("w2",paire[2]);
		    w2=paire[2];
		    break;
		  case 'g':
		    //console.log("w3",paire[2]);
		    w3=paire[2];
		    break;
		  case 'h':
		    //console.log("da",paire[2]);
		    da=paire[2];
		    break;
		  case 'i':
		    //console.log("ts",paire[2]);
		    ts=paire[2];
		    break;
		  case 'j':
		    //console.log("ga",paire[2]);
		    ga=paire[2];
		    break;
		  case 'k':
		    //console.log("ba",paire[2]);
		    ba=paire[2];
		    break;
		  case 'l':
		   // console.log("ar",paire[2]);
		    ar=paire[2];
		    break;
		  case 'm':
		   // console.log("an",paire[2]);
		    an=paire[2];
		    break;
		  case 'n':
		    //console.log("it",paire[2]);
		    it=paire[2];
		    break;
		  case 'o':
		    //console.log("wr",paire[2]);
		    wr=paire[2];
		    break;
		  
		}
		//resultats.push(paire);
	}
	
  return {
    te:te,
    pr:pr,
    hu:hu,
    lu:lu,
    w1:w1,
    w2:w2,
    w3:w3,
    da:da,
    ts:ts,
    ga:ga,
    ba:ba,
    ar:ar,
    an:an,
    it:it,
    wr:wr,
  }
  
}

That can be a source of error as well?