Is RSSI different for each BLE channel?

[suke]

Hello.

I have two nRF52832 (central and peripheral).
Connect the central and peripheral and call sd_ble_gap_rssi_get() on the central to get the RSSI and channel.
The attached photo is a collection of data collected for a certain period of time.

From my point of view, RSSI is the receiving sensitivity, and if the distance to the peripheral and the transmission power of the peripheral are the same, I think that RSSI is the same for all channels. However, in the results, there is a difference in RSSI for each channel. When we checked the wi-fi analyzer in the evaluation environment, we confirmed that wi-fi existed in the frequency band where RSSI was weak.
Below are the questions.

1. When radio waves interfere (wi-fi exists nearby), RSSI weakens?

2. Can RSSI be acquired even if BLE packet cannot be received normally (ex.CRC error)?

3. I want to make sure that the BLE connection status is maintained for a certain time or longer.
Do you have any reference material?

Thank you.

[haakonsh]

From my point of view, RSSI is the receiving sensitivity, and if the distance to the peripheral and the transmission power of the peripheral are the same, I think that RSSI is the same for all channels.

That's not true. 

The Received Signal Strength Indicator is a measure of the amount of RF energy that was received at the time of sampling. Any RF source in the receivers frequency band will affect the RSSI value. The RSSI is sampled by and ADC in the receiver, it cannot distinguish between different RF sources. 

Also, a receiver's sensitivity is a measure of a receiver's ability to distinguish a signal from the noise floor, ie. how weak a signal can be and still be received accurately. 

1. No, it increases. 
   
   
2. Yes.
   
   
3. Can you elaborate a bit more on what it is you're trying to achieve?

[suke]

haakonsh-san
Thank you for your reply.

Results show the RSSI is low in the specific frequency band (17-19, 30ch). If not because of interference with wi-fi, what else could be the cause?

I want to show the communication disconnection rate for each distance between the central device and peripheral device.
I tried it outdoors in a large area with no surrounding buildings, but the results differ depending on the day. (Perhaps the influence of radio wave reflection, humidity, etc...?)
Therefore, set it in a good environment such as an anechoic chamber to obtain reproducible results. Then, by reducing the transmission power, I thought that it would be possible to reproduce the state where the distance was long (weak RSSI).
In other words, it is not the communication disconnection rate for each distance, but the communication disconnection rate for each RSSI value. I thought that if I knew the communication disconnection rate for each RSSI, I could refer to how much communication stability I could expect depending on the installation environment of the device.
However, I heard that communication stability requires checking not only RSSI but also data throughput. I would like to know if there is an appropriate method for making such an evaluation.
Thank you very much.

[haakonsh]

suke said:

Results show the RSSI is low in the specific frequency band (17-19, 30ch). If not because of interference with wi-fi, what else could be the cause?

It could be the absence of wifi in those channels that makes them show a lower RSSI. Note that any 2.4GHz RF source will be added to the RSSI, like zigbee, ANT, 802.15.4, other proprietary protocols, and even a microwave oven. 

suke said:

I tried it outdoors in a large area with no surrounding buildings, but the results differ depending on the day. (Perhaps the influence of radio wave reflection, humidity, etc...?)

 Exactly! 

suke said:

In other words, it is not the communication disconnection rate for each distance, but the communication disconnection rate for each RSSI value. I thought that if I knew the communication disconnection rate for each RSSI, I could refer to how much communication stability I could expect depending on the installation environment of the device.
However, I heard that communication stability requires checking not only RSSI but also data throughput.

 RSSI is really not a good representation of Link Quality, the throughput is a much better metric. 

Also, note that an antenna has a different impedance match throughout the frequency-band it's tuned for. See this plot of the Standing Wave Ratio, a measure of the amount of energy that is reflected back from the antenna into the transmitter(not transmitted):

This plot shows an ideal impedance match at 2440MHz with an exponentially decreasing impedance match near the ends of the 2.4GHz frequency band. The RSSI values with this antenna would be greatest around 2440MHz and decrease near the edges. This plot is only valid for the given antenna that was measured, your antenna will have a different characteristic with respect to the impedance match. 

suke said:

I would like to know if there is an appropriate method for making such an evaluation.
Thank you very much.

I suggest you use the Radio Transmitter Example and the Radio Receiver Example (no SoftDevice), or the Experimental: ATT_MTU Throughput Example (with SoftDevice). 
The Throughput example already measures throughput. 

[suke]

Thank you for the detailed information.
I received a very satisfactory answer !!!