NRF9160 Zephyr ncs ZTE-M/NB-IOT Multiple Cell-Tower RSSI

[Thomas Li Fredriksen]

Good afternoon,

I have previously asked the following question ("NRF9160 Zephyr ncs ZTE-M/NB-IOT Cell-Tower RSSI"):

https://devzone.nordicsemi.com/f/nordic-q-a/42241/nrf9160-zephyr-ncs-zte-m-nb-iot-cell-tower-rssi

Unfortunately, I was a bit quick to accept the answer, as the suggested solution only seem to provide the RSSI for a single tower.

I am looking for some way of reading out the RSSI of multiple towers, preferably all towers within range.

Is this at all possible?

/Tom

[Thomas Li Fredriksen]

I did find the command "AT%NBRGRSRP", which seems promising, but I cannot seem to make sense of the provided Cell-Id, which appears to be only two or three digits?

[Håkon Alseth]

All commands that has a "AT%" are notification subscriber commands, so you should get a notification back if the subscribed command has a notification for you. Enabling it should not have a field for cell-id, but the returned data from the modem will. Could you post what you get back?

Cheers,

Håkon

[Thomas Li Fredriksen]

Hi Håkon,

This is what I get when running "AT%NBRGRSRP" in the LTE-monitor:

AT%NBRGRSRP
%NBRGRSRP: 157,6400,53,170,6400,46,168,6400,37,156,6400,33
OK

/Tom

[Håkon Alseth]

Hi Tom,

Thomas Li Fredriksen said:

%NBRGRSRP: 157,6400,53,170,6400,46,168,6400,37,156,6400,33

I see where you're confused (I was too). This provides the physical cell id (PCI), and not the cell identity.

157 = Physical Cell identity (http://lteuniversity.com/ask_the_expert/f/69/t/589.aspx)

6400 = EARFCN (https://www.cablefree.net/wirelesstechnology/4glte/lte-carrier-frequency-earfcn/)

53 = RSRP (-140 + 53 = -87 dbm)

 

Best regards,

Håkon

[Thomas Li Fredriksen]

Hi Håkon,

This appears to make more sense, however it does not resolve my problem.

I still need some way of reading the cell ID, not the PCI. I assumed this would be achieved by some AT-command, however some SDK-way of achieving this would be preferable.

/Tom

[Håkon Alseth]

Hi Tom,

 

You cannot get the cell id from neighboring cells via the AT commands, but you will get the PCI and RSRP.

AFAIK, it is technically possible to decode the Cell ID from neighboring cells, but we do not support it, and its time and power consuming. Is this a hard requirement for your application?

Kind regards,

Håkon

[Thomas Li Fredriksen]

Hi Håkon,

I am experimenting with non-GNSS location and tracking, and need the location of all towers within range.

From what I understand, I do need the cell identifier to get this. However, might it be possible to look up the cell ID from the PCI given I have the cell ID of at least one tower?

Secondly, on a related note, do the NRF9160 support any alternative non-GNSS tracking solutions?

/Tom

[Håkon Alseth]

Hi Tom.

 

Thomas Li Fredriksen said:

From what I understand, I do need the cell identifier to get this. However, might it be possible to look up the cell ID from the PCI given I have the cell ID of at least one tower?

I am not sure that it is possible, as the pci is a number between 0 - 517, so not unique at all in the bigger picture.

Thomas Li Fredriksen said:

Secondly, on a related note, do the NRF9160 support any alternative non-GNSS tracking solutions?

 No, there's no other tracking solutions in the chip apart from the gps and modem itself. You can find external ICs that communicate over SPI/I2C that might give tracking capabilities, but this is a field you'll have to research. 

Kind regards,

Håkon

[Sashi]

Is there any plans to support cell ID scanning or is there another command that can be used for tower location capabilities. For our application this is a hard requirement, and the additional time and power consumption is acceptable. 

[Håkon Alseth]

If there are plans for this: That is not something I can answer directly, as I do not know. I would recommend that you contact your local sales manager and ask about that feature. If you do not have the contact details of your local sales manager, please send me a direct message with your location.

Do also note that you are posting in a quite old case (9 months), which can be problematic for the engineers in support to pick up on. Next time, please create a new ticket, and rather link to the older one.

 

Kind regards,

Håkon