Explicit feedback with sounding packet May 2006 doc.: IEEE 802.11-06/0xxxr0 July 2006 Explicit feedback with sounding packet Date: 2006-07-17 Authors: Notice: This document has been prepared to assist IEEE 802.11. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE 802.11. Patent Policy and Procedures: The contributor is familiar with the IEEE 802 Patent Policy and Procedures <http:// ieee802.org/guides/bylaws/sb-bylaws.pdf>, including the statement "IEEE standards may include the known use of patent(s), including patent applications, provided the IEEE receives assurance from the patent holder or applicant with respect to patents essential for compliance with both mandatory and optional portions of the standard." Early disclosure to the Working Group of patent information that might be relevant to the standard is essential to reduce the possibility for delays in the development process and increase the likelihood that the draft publication will be approved for publication. Please notify the Chair <stuart.kerry@philips.com> as early as possible, in written or electronic form, if patented technology (or technology under patent application) might be incorporated into a draft standard being developed within the IEEE 802.11 Working Group. If you have questions, contact the IEEE Patent Committee Administrator at <patcom@ieee.org>. Joonsuk Kim, Broadcom Matthew Fischer, Broadcom

CIDs to cover in LB84 comments July 2006 CIDs to cover in LB84 comments CID numbers to cover 11959, 947, 948, 949, 950, 951, 4144, 1072, 952, 4145, 10286, 954, 4147, 957, 4150, 969, 1073, 4161, 968, 970, 4526, 2697 Joonsuk Kim, Broadcom

July 2006 Current draft D1.0 Current draft D1.0 has conflicts in PHY and MAC sections In 9.20.3.3 (Page 124, line 15), the sequence is initiated by an AP which sends an un-steered sounding PPDU for CSI feedback request. We don’t have this issue anymore with 06/886r2 accepted. In 20.3.6.3 (Page 221, Line 16), when using a spatial mapping matrix Qk , it needs an estimate of the effective channel HkQk , where Qk is in this case a NtxxNSTS orthonormal spatial mapping matrix. The sounding PPDU sounds the effective channel that includes spatial mapping matrix Qk at the transmitter. Joonsuk Kim, Broadcom

July 2006 Definitions H_eff: a transfer function from the transmitter spatial mapper input to the receiver FFT outputs (CID 3962) H_eff = HQ Q: Spatial mapping matrix used at beamformer It can be either direct mapping, spatial expansion or beamforming steering matrix Q_steer: Steering matrix used for beamformed packet transmission (special name of Q only if Q is used for steering matrix) Q_steer = QV V: Beamforming feedback matrix to be fedback from beamformee to beamformer Joonsuk Kim, Broadcom

The explicit feedback with current sounding packet July 2006 The explicit feedback with current sounding packet We may have a spatial mapping matrix, Q, for sounding packet in the explicit feedback Beamformee may estimate the channel, where the measured channel includes Q, to feedback either the full effective channel or beamforming matrix V. For example of SVD BF, HQ = USV* Joonsuk Kim, Broadcom

July 2006 Proposed solution Beamformee may return HQ for full channel feedback, or V for beamforming matrix feedback. CSI feedback When Beamformee sends back the measured channel, the measured channel includes the Q matrix which is used in the sounding packet. We should define H_eff = HQ, which is the effective channel including the spatial mapping matrix (CID 11959, 952, 4145, 10286, 3962,11979) The beamformee removes the CSD in table n67 from the measured channel matrix before it returns CSI feedback with H_eff. Joonsuk Kim, Broadcom

Proposed solution (cont.) July 2006 Proposed solution (cont.) V matrix feedback (either compressed or uncompressed) The beamformee removes the CSD in table n67 from the measured channel matrix before it determines the beamforming feedback matrix V. The beamformee sends back V and Beamformer needs to modify it to account for Q which is originally used in the sounding packet. We introduce Q_steer. (CID 947, 948, 949, 950, 951, 4144, 1072, 954, 4147, 957, 4150, 969, 1073, 4161, 968, 970) The beamformer shall use Q_steer = QV for new steering matrix by multiplying fedback V with the original Q used in the sounding packet. H_eff = HQ = USV* for the example of SVD operation HQQ*QV = USV*Q*QV  HQV = HQ_steer = US Joonsuk Kim, Broadcom

Conclusions We can resolve it by introducing H_eff and Q_steer. July 2006 Conclusions We can resolve it by introducing H_eff and Q_steer. H_eff includes Q but not Ccsd in frequency domain. Q_steer is a special name of Q, when Q is used as a beamforming steering matrix. Beamforming matrix V is found from H_eff for the feedback, and Q_steer = QV is new steering matrix for beamformed packets. Normative text is provided in 06/1335 Joonsuk Kim, Broadcom

Question to ask at the adhoc group July 2006 Question to ask at the adhoc group Measured channel is HQCcsd , where Ccsd is taken out at the beamformee before the beamformee sends the feedback either Heff, uncompressed or compressed V matrix, where Heff = HQ and V is found from Heff. Otherwise, the beamformer should consider Ccsd from the measured channel before it finds the new steering matrix. So, it is a matter of choice, whether the action is taken at the beamformee or the beamformer. When the beamformee takes it out Heff = HQ, Q_steer = QV When the beamformer takes it out Heff = HQCcsd, Q_steer = QCcsdV Joonsuk Kim, Broadcom

Advantages when Ccsd is taken care of at the beamformee July 2006 Advantages when Ccsd is taken care of at the beamformee The steering matrix block, Spatial Mapping Q in Figure n51, is the place where the packet is beamformed. Conceptually, it makes much sense to define the channel, which the steering matrix is found from. Ccsd is a phase-shift matrix which can be easily handled by the beamformee. Joonsuk Kim, Broadcom

July 2006 Straw Poll Who do you think needs to take out Ccsd from the measured channel (Heff is the channel to be fedback from the beamformee)? *NOTE: Ccsd refers to the CDD block located between STBC and Spatial Mapping blocks in Figure n51 (D1.03) The beamformee (i.e., Heff = HQ, Q_steer = QV) Votes: The beamformer (i.e., Heff = HQCcsd, Q_steer = QCcsdV) Joonsuk Kim, Broadcom

July 2006 References 11-06-0675-035-000n-tgn-d1-0-lb84-beamforming-and-adapation.xls Joonsuk Kim, Broadcom