Efficient Frame Format for MU Transmission January 2010 doc.: IEEE 802.11-10/0084r3 Efficient Frame Format for MU Transmission Date: 2010-11-08 Authors: Brian Hart, Cisco Systems

Abstract MU-MIMO is a key technique to attain at least 1 Gbps throughput at MAC SAP for Multi-STA. But, the gain of MU-MIMO maybe shrink significantly due to mismatching of the lengths of co-transmission STA’s PSDU. A sustainable frame format is proposed here to exploit the gain of MU-MIMO against the mismatching problem for co-transmission STAs

The mismatching problem There exists a mismatching problem for MU-MIMO transmission for 802.11ac due to mismatching the lengths of co-transmission STA’s PSDU. An example is given here: STA1’s transmission rate is 500Mbps and 400Mbps for SU-MIMO and MU-MIMO transmission, respectively STA2’s transmission rate is assumed as same as that of STA1 for simple. PSDU data length of STA1 is 2000 kb, therefore 5ms for MU-MIMO transmission PSDU data length of STA2 is 800 kb, therefore 2ms for MU-MIMO transmission STA 1 STA1-PSDU STA 2 STA2-PSDU

Transmission schemes for Multi-STA with mismatching lengths of their PSDU Direct transmission utilizing MU-MIMO The LENGTH field in L-SIG is set according the maximum among the lengths of these multi-STA’s PSDU The LENGTH field in each STA’s VHT-SIG-B is set according the actual length of the STA’s PSDU The transmission duration for the example is 5ms Direct transmission utilizing SU-MIMO for each STA in turn The transmission duration for the example is 5.6ms without accounting for the overheads of preamble, SIFS, etc. STA1-PSDU STA2-PSDU

An efficient transmission scheme based on a new frame format A new frame format with sustainable VHT-segment proposed here: A series of VHT-segments with the maximum of 4 can be concatenated together to form the whole frame. Each VHT-segment consists of VHT-SIG-A, VHT-STF, VHT-LTFs, VHT-SIG-B and VHT data

An efficient transmission scheme based on a new frame format (con’t) A new frame format with sustainable VHT-segment proposed here: The LENGTH field in L-SIG is set according to the total length of the whole frame consisted of all VHT-segments

An efficient transmission scheme based on a new frame format (con’t) A new frame format with sustainable VHT-segment proposed here: Each STA’s NSTS field in VHT-SIG-A implies whether its PSDU has been ended or not

An efficient transmission scheme based on a new frame format (con’t) A new frame format with sustainable VHT-segment proposed here: The length field in VHT-SIG-B of each VHT-segment indicates the actual length of VHT-data in current VHT-segment.

Gain and Conclusion The transmission duration for the former example is 4.4ms, therefore a gain of more than 10% can be obtained. Although the transmission scheme based on the proposed frame format is somewhat like MU-TXOP, but it is superior to MU-TXOP due to omitting SIFS, L-STF and L-LTF. Nevertheless, TXOP/MU-TXOP isn’t substitutable in providing continuous transmission with longer duration (for example more than 5.4ms).

References Specification Framework for TGac, IEEE 802.11-09/0992r15, September 2010