There is a special procedure called sounding procedure in 802.11ac (please see the attachment). This procedure helps to realize transmit beamforming. Beamformer (transmitter-most probably AP) should create beamforming matrix to send its data to beamformee (receiver-most probably station).
Procedure is as follows:
Upon association (where AP assigns 16 bits association ID -AID- to the user),
1) AP first sends NDP-A packet (this is MAC packet, it includes station's AID info, so station can be identified by AID, not MAC address),
2) AP sends a NDP packet (this is PHY packet, not MAC packet. It also contains intended user's AID, so user recognizes this packet. The reason of using this PHY packet is to determine channel parameters between AP and station. PHY packet is used here to reduce overhead, because network does not use MAC packet here).
3) Station receives PHY packet, estimates the channel parameters (by using this PHY packet that is NDP packet), and feedbacks it to AP with Compressed Beamforming packet (this is also MAC packet).
4) Now, since AP knows the channel parameters, it can create its beamforming matrix for the transmission.
Q: MAC address is 48 bits. However, this procedure uses 16 bits AID to identify the user. Do you think is it okay for the network to use this 16 bits PHY layer AID instead of 48 bits MAC address? Is there here a contention problem or something?
- Badges
- Science topic
- Similar topics
- Engineering
- Telecommunications Engineering