RX quality differs significantly between devices: RSSI / SNR

I am having troubles to understand the following:

I am installing two devices (Dragino LT-33222-L with an external antenna and Elvaco CMi4110) in a building located around 500m from the next Gateway (Mikrotik WAP LR8 Kit). It is not line of sight, but there are just a few walls in between.

When using an adeunis network tester right next to both devices the RX strength is as I would expect:

  • SF: 8
  • RSSI: ~ -110
  • SNR: ~ 0

However, both above mentioned devices are received significantly worse by the gateway:

  • SF: 10-12
  • RSSI: ~ -110
  • SNR: -15 - -20

While the RSSI is roughly the same, the SNR is significantly worse.

What could be the issue here? Any idea?

Please see this post. Also, in LoRaWAN, not having sufficient separation/distance between sensor & gateway is a problem - this video shows how to measure this.

In my experience, even given a challenging environment (lots of metal walls, many layers of rebar-concrete, long distances, heavy EMI, industrial manufacturing facility, etc), when one uses only quality OEM antennas from proven vendor (Laird, Semtech, Murata, etc), LoRaWAN is so incredibly robust that two gateways are plenty (really for redundancy purposes, as one is on WiFi and the other wired Ethernet :slight_smile: ).

I was spooked reading about poor RSSI/SNR and assumed the worst, so procured six gateways. but have installed only four and I find that my link budget is very good even in worst-case. So, I am planning to scale it down to three and keep the fourth installed but off, as a ready-spare.