Is LBT (Listen-Before-Talk) supported by The Things Gateway?


(Hoon Park) #1

Is LBT (Listen-Before-Talk) supported by The Things Gateway?

For countries like Korea, LBT is required. All devices must use LBT. It seems like sx1301 + sx1272 chip + FPGA are required to support LBT on a gateway.

I'd like to know if The Things Gateway supports LBT.


Single Channel Gateway part 1
LoRa Initial Transmission Phase?
(Richard V) #2

To my knowledge Lora devices are not listening before they transmit.


(Arjan) #3

For countries like Korea, LBT is required. All devices must use LBT.

I think it's not mandatory in the protocol; here is what I once found on the subject for nodes: Do LoRa(WAN) nodes perform any activity detection and/or collision detection?


(Hoon Park) #4

@arjanvanb, I've read your comment before I posted my question on the forum. LoRaWAN may not mandate LBT but Korean regulation mandates LBT. The current KR920-923 ISM band regulation exclusively uses LBT to maximize the MACPayload.

It's why I need LBT support if I wanted to use any device (including gateways) in Korea.


#5

https://www.thethingsnetwork.org/forum/t/is-the-things-gateway-open-source/3423/5

When they are ready to ship, we will know its capabilities :slight_smile:


#6

I wondering about "use LBT to maximize the MACPayload".

Regardless of the size of the data i want to send, LBT is just listening channel before data sent.

Q1. If i use LBT and check out channel is free, can i sent maximum length of payload ??

Q2. At this time, Why LoRaWAN devices does not use LBT or not implemented yet??.
In wireless sensor network, many nodes use LBT or CSMA/CA to avoid collision which has power consumption but do that.


#7

In which region? And how are you going to meet RX windows timing requirement?

This functionality was removed from Semtech reference sources more than year ago. First, it turns LoRa from Long Range to Short Range. Second, it's almost useless due to "hidden node problem".


#8

@hobo Thank you for reply.

i have another question about RX windows and gateway.

Q1. After Node transmit(Tx), gateway will transmit ACK or data(If it presence) to RX1 window. But if there is a collision at RX1 window, how gateway know this and the gateway sends data to RX2 window?


(Arjan) #9

It doesn't know.

  • It might be able to detect a collision, but even if no collision occurred then it would not know if the node received the data.
  • It's not like RX2 is used when RX1 fails: just one window is selected by the backend (it's not up to a gateway to decide), and TTN prefers RX2 so probably skips RX1 altogether.

So, if a node requests an ACK for an uplink but does not receive it, or if an application requests an ACK for a downlink but does not get it, then they need to decide for themselves to try again.

(Note that ACKs are very limited anyhow.)


#10

@arjanvanb Oh, Sorry about my mistake, Thanks.

and then,,,

Q1. How Rx slot size is decided? is there any default value?

Q2. If there isn't any preamble in Rx2, how retransmission is decided in 1~3 seconds?


(Arjan) #11

....time to read the specifications, in which you'll also run into "18.1 Uplink Timing Diagram for Confirmed Data Messages".


LoRa Maximum Transmission Unit
(Gmbankar) #12

Is it possible to implement LBT in IN865 Band devices?