DECT NR+ MAC->PHY transmission scheduling

Hi there,

One of the functions of the DECT MAC layer is to allocate radio resources to RDs.  According to the standard, the MAC layer does that using special beacon messages ("resource allocation IEs") that contain the slot and subslot numbers that are assigned to each RD.

I'm trying to develop a simple MAC layer on top of the PHY. I don't quite understand how the MAC layer can instruct the PHY to use only certain slots or subslots in the transmission. The only relevant parameter I see is the ability to set the start_time when calling nrf_modem_dect_phy_tx(). But that is set in "modem time" rather than slots. Is the MAC layer responsible for syncing the timing of frames/slots?? Shouldn't that capability be provided by the modem? (I'm a newbie, so hopefully what I'm asking makes sense).

Secondly, , there's a binary demo you're providing for the nrf9161 with a Mosh shell and a "dect" command that implements, among other things, a (partial?) MAC layer. Will this source code become public, or be provided as a library?

Thanks,

Nir

Parents
  • Hi,
    You are correct, the MAC layer has to allocate and control radio resources. It does not however need to be per RD, but instead the FT device announces the random access channel resource, and the PT devices compete for access on the slots, using the Listen-Before-Talk LBT to detect if somebody is already transmitting.
    MAC layer needs to translate modem time to slots. When the beacon message is received, it is the start of a frame, received with the modem timestamp. MAC layer can then calculate the timestamps for slots based on the timestamp. As said, usually it is not per slot, but there is a longer random access window of multiple slots.
    the binary demo is currently not planned to be released as source code, it is a SW tool, not a reference implementation of MAC.

    BR Lauri  

Reply
  • Hi,
    You are correct, the MAC layer has to allocate and control radio resources. It does not however need to be per RD, but instead the FT device announces the random access channel resource, and the PT devices compete for access on the slots, using the Listen-Before-Talk LBT to detect if somebody is already transmitting.
    MAC layer needs to translate modem time to slots. When the beacon message is received, it is the start of a frame, received with the modem timestamp. MAC layer can then calculate the timestamps for slots based on the timestamp. As said, usually it is not per slot, but there is a longer random access window of multiple slots.
    the binary demo is currently not planned to be released as source code, it is a SW tool, not a reference implementation of MAC.

    BR Lauri  

Children
No Data
Related