Closed drasko closed 7 years ago
It integrates a sx1276 transceiver via SPI. The RPI can be connected to 2 SPI devices, so this gateway can handle at most 2 concurrent clients transmitting at the same time on different LoRa modulation parameters. But with TDMA it can handle more than 2 clients.
The gateway uses LoRa (SX1276) but does not use nor is it compatible with LoRaWAN. It's a specific solution for tracking high altitude balloons which run matching software (https://github.com/piinthesky or https://github.com/daveake/FlexTrack).
Re number of clients, the limit is not 2. Many more can be used with TDMA.
This TDMA can be implemented in SX1276 FW or in RPi SW, as a LInux app (or driver)?
Yes, both clients I linked to support this (AVR/Arduino and C/Linux)
So - TDMA must be implemented on both client and GW side, and you must have some kind of sync reference...
On the GW side - do you implement TDMA in SX1276 FW or directly in Linux on RPi?
Yes, on both sides, using GPS timing. Every balloon tracker has GPS anyway.
Everything is done as C applications running on the host.
I can not see from here: http://www.hoperf.com/rf/lora/RFM98W.htm which Semtech transceiver this module integrates. I guess it is a custom chip with Semtech IP core.
Anyway, as I understand, for LoRa GW we need:
How come that this soultion does not need Concatenator and/or I/Q modulator?
How many clinets (end-nodes) can be connected to a GW like this?