Observed sticker rssi values


In my project I am using raspberry pi as receivers for estimote stickers. I have questions about the rssi values I am receiving. It almost seems like the sticker is emitting at two transmission powers. In the attached time series plot the epoch time is along the x axis and the rssi values are on the y axis. This experiment has two stickers and one pi. The results are surprising. I expected the noise, but I did not expect to see what looks like two signals for each sticker.

Has anyone else observed something similar?

The results in the picture are in a conference room on a table without many objects for interference. I have observed similar results for other experimental configurations.

I am broadcasting the nearable packet at -12 power with an interval of 1000 mHz for each sticker. The stickers are stationary for the duration of the experiment.


That is quite interesting. Stickers should advertise with the same power for each packet. However I have some ideas:

  1. There might an obstacle that reflects packets. This would mean that some of the packets travel directly and some gets reflected by obstacle that absorbs some of the radio wave energy. RSSI chart looks like one sticker is reflecting better that the other. Try to change their position (or swap places) and see if pattern repeats.

  2. I suspect this may be caused by the fact that BLE advertises on three different channels (37,38,40) that are spaced in the lower, middle and upper part of BLE assigned radio band. Receiving and transmitting antennas might have different gain on different channels. Since BLE advertising cycles though those channels, sometimes RSSI is higher, sometimes lower. Nearable packet contains information about channel, so you can check that.

1 Like