Low RF RSSI on Duty Cycled Reception

[Shaunus]

Good day,

I am currently working on the development of an IoT sensor node product based around the STM32WLE5CC MCU.

The product needs to be able to wake up from sleep on reception of a LoRa RF message. I have attempted to accomplish this by transmitting a long preamble and having the node receive the message by setting the radio in RxDutyCycle mode prior to sleeping. The timing of my duty cycle is such that I am guaranteed an Rx window within the preamble duration. This wakeup functionality works, however, the RSSI of the received wakeup signal is shown to be >20dBm lower than if I had just used continuous reception. This is a problem as we are now encountering suspected range issues where units can detect the preamble but fail to fully receive the associated payload.

I have gone through multiple test setups with different spreading factors and the results are consistent across the tests:

Test 1.1 (SF = 7, BW = 125 kHz, Preamble Len = 500, Rx = Full Receive)

• Wake RSSI = -44
• Wake SNR = 12

Test 1.2 (SF = 7, BW = 125 kHz, Preamble Len = 500, Rx = Duty Cycle [Rx: 5 ms, Sleep: 495 ms])

• Wake RSSI = -65
• Wake SNR = 12

Test 2.1 (SF = 10, BW = 125 kHz, Preamble Len = 64, Rx = Full Receive)

• Wake RSSI = -42
• Wake SNR = 9

Test 2.2 (SF = 10, BW = 125 kHz, Preamble Len = 64, Rx = Duty Cycle [Rx: 45 ms, Sleep: 455 ms])

• Wake RSSI = -66
• Wake SNR = 4

Test 3.1 (SF = 12, BW = 125 kHz, Preamble Len = 16, Rx = Full Receive)

• Wake RSSI = -44
• Wake SNR = 7

Test 3.2 (SF = 12, BW = 125 kHz, Preamble Len = 16, Rx = Duty Cycle [Rx: 165 ms, Sleep: 335 ms])

• Wake RSSI = -65
• Wake SNR = 4

The only indication I can see of some discrepancy is that the full reception results in only a radio interrupt IRQ 0x02 (Rx Complete), where as the duty cycled reception results in an IRQ 0x04 (Preamble Detected) followed by the IRQ 0x02. I have attached the debug logs for the SF = 7 tests.

I would really appreciate any assistance or explanation for these findings as this is a core functionality of the system and is causing reliability issues with our field tests.

Many thanks,

Shaun

[Filippo_Malleo]

Hi Shaunus,

it seems that the duty cycle settings are too challenging: RX window too small to have an accurate RSSI value. 
Try increasing the RX window to get more energy from the incoming packet and check that both RSSI and SNR values increase.

BR,
Filippo