So here’s a question that’s not likely asked often. I’m testing the Boron LTE and having good results with it so far. I use a third party SIM from Hologram since my customers often live in areas of the US that AT&T doesn’t reach reliably. So far I have the Boron successfully connecting to both Verizon and AT&T, and it even seems to be automatically selecting the carrier. That’s great!
BUT. My last batch of prototypes (not using Particle hardware) were supposed to intelligently switch between the two carriers also. I found out it wasn’t so intelligent and ended up with devices stuck in endless connection loops trying to connect to a carrier that it simply couldn’t reach.
For testing purposes, I’d like to simulate some of the conditions I actually see in the field, like carrier dead zones. I have locations where AT&T is either unavailable or has extremely low signal, but Verizon works fine. My question is, does anyone have ideas on how to simulate this environment in the lab? i.e. One carrier is accessible, but the other is not. I can put some money towards a simulation environment, but I don’t exactly have a big budget. Is there a decent way to for instance, block carrier X or degrade its signal, on a very localized scale? What I am envisioning and may be completely off base with, is being able to jam a particular carrier’s signal, just within my lab, so I can see how the device performs.
For clarity, are your Particle Boron devices having trouble or just your other ones?
Realistically, the only thing I can think of would be to use a fairly large Faraday cage, and then use a Verizon mini cell tower cell hotspot to generate a local cell signal. EDIT: that said, it’s likely that cell tower hotspot wouldn’t be compatible with CAT M1, so not super likely.
Alternatively, if you just get a Verizon only SIM card from Verizon for testing purposes, that may work?
Also, I’ve seen some similar things posted before, and there was a recommendation to manually set providers on the modem, depending on which you are trying to access, or maybe a timeout condition.
Doing any kind of active RF suppression in the air would be very illegal, probably. You could run some in-line notch filters after your antenna if you could identify the AT&T frequencies. Just get an adapter to the standard coax size and then get some inline coax filters or make an analog filter circuit test board.
PS: How has it been with the Boron LTE and Hologram? I use Hologram with our 3G Electrons, but was not able to get it to work with the Boron LTE in the lab. Do you have any tips or interesting things you’ve discovered?
My Particle devices so far are behaving well. But so were the others until I put them into the field. What I’m looking to do is simulate conditions where either one or the other carrier is unavailable or is unreliable. That way I can tweak my code to attach to the best available, or best case, show that the SIM and Hologram are able to select the best available carrier upon connect.
I’m still experimenting with settings, but so far I’ve been able to get reliable connections via Hologram to both AT&T and Verizon, and can select in code which to connect to. More testing is required to see if this works consistently. Just today I’ve been able to get the Boron w/ Hologram SIM to connect to both carriers without code changes. What I need to test though is the field conditions these devices will be placed in. My prototypes are over 500 miles away, so I have only infrequent physical access to them.
I hadn’t thought about using notch filters on the antenna line. That sounds interesting and is something I’ll look into further. I’m not opposed either to some type of signal interference to simulate conditions, assuming I can localize it to a very small space near the unit under test. The notch filter idea though seems like a better solution though.
As to how I connect to Hologram, I am still trying different settings. So far a few things have worked well, but I’m still early in my testing.
Connecting to Hologram:
Set APN to “hologram”
Hologram with Verizon:
AT+COPS=2
AT+URAT=7
AT+UMNOPROF=3
AT+CFUN=15
Connecting to Hologram with AT&T:
AT+COPS=2
AT+URAT=7
AT+UMNOPROF=0
AT+CFUN=15
The connections take a relatively long time (50-120 seconds) but have been pretty solid so far. I suspect the long delay is due to the SIM reset performed with AT+CFUN=15.
Just today I replaced all of those commands with simply AT+UMNOPROF=0 and Cellular.connect(). That seems to be auto-switching between AT&T and Verizon. More testing needed on that. Also, I’m not sure which of the previous AT commands I’ve used are stored in NVRAM and are thus affecting this latest test.
Have you tried any of the network providers to see if they have test labs? Here in the UK and elsewhere in Europe (but not much help for you based in the US!) Vodafone have a number of labs where you can go and test equipment. They are more focused on NB-IOT rather than LTE-M now but I understand the other networks EE and O2 are starting rollout. It is tremendously difficult otherwise to subject devices to all the possible signal conditions they might encounter. I know this from WiFi device testing and at least there you can more easily recreate conditions.