1 / 13

MCS Control of Sounding Feedback for Feedback Overhead Reduction

MCS Control of Sounding Feedback for Feedback Overhead Reduction. Authors:. Date: 2010-11-09. Abstract. Explicit sounding feedback overhead is a major issue in MU mode effectiveness Using higher MCSs for sounding feedback can dramatically reduce overhead compared with lowest MCS

Télécharger la présentation

MCS Control of Sounding Feedback for Feedback Overhead Reduction

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. MCS Control of Sounding Feedback for Feedback Overhead Reduction Authors: Date: 2010-11-09 Nir Shapira et al, Celeno Communications

  2. Abstract • Explicit sounding feedback overhead is a major issue in MU mode effectiveness • Using higher MCSs for sounding feedback can dramatically reduce overhead compared with lowest MCS • We present simulation results for different aging scenarios • We suggest the beamformer to recommend an MCS for the feedback frame for each user using the FLA mechanism • We suggest to separate the MCS recommendation for traffic and sounding feedback Nir Shapira et al, Celeno Communications

  3. Simulation Setup • We show the effect of sounding feedback overhead in various sounding frequencies • We compare overhead in lowest MCS and overhead with “optimal+safety margin” MCS selection • We limit feedback frame MCS to 1SS, no BF • Simulation calculates throughput assuming worst case aging (i.e. all frames are equally aged according to sounding frequency) • We assume CV feedback with b_phi=8, b_psi=6 [1] • We show results, using Ng=2, for two cases: • 8TX beamformer and four 3RX users (8,3,3,3,3) • 4TX beamformer and four 2RX users (4,2,2,2,2) Nir Shapira et al, Celeno Communications

  4. Net Aging Effect:Simulation Results w/o Feedback Overhead • Curves ordered, as expected, according to aging Nir Shapira et al, Celeno Communications

  5. Results with Overheads (Ng=2), (8,3,3,3,3) • At lowest rate feedback, short sounding period is highly degraded • “Optimal” feedback MCS closes most of the throughput gap to no-overhead case • With MCS selection, best sounding period is 10-20 ms Nir Shapira et al, Celeno Communications

  6. Results with Overheads (Ng=2), (4,2,2,2,2) • “Optimal” feedback MCS closes most of the throughput gap to no-overhead case • With MCS selection, best sounding period is 10 ms Nir Shapira et al, Celeno Communications

  7. Why Beamformer should Control MCS of Feedback • MU Beamformer should be able to predict overhead for grouping and scheduling decisions • Uplink traffic might be sparse • Not enough statistics for users to determine MCS according to PER • Does not justify sending uplink sounding frames • Users have too many unknowns in link budget estimation • AP total power, implementation loss, noise figure • In MU operation AP has updated uplink channel estimate from previous sounding exchange • Can give good MCS recommendation with minimum SNR margin Nir Shapira et al, Celeno Communications

  8. The need to separate MCS recommendation for traffic and for sounding feedback • It makes sense to use the already defined FLA mechanism for controlling MCS of sounding feedback • Different QoS requirements for sounding feedback and traffic • Uplink traffic might use different transmit technique (might be beamformed, in SU or MU) • Beamformer might not want to recommend MCS for uplink traffic • Might be unaware to uplink traffic QoS needs, frame/aggregation size, etc • Might not have full channel knowledge in case no uplink sounding is available • FLA recommendations incur minimum overhead • No significant overhead in case different recommendations used for sounding feedback and traffic • Separation of MCS recommendation will increase the chance of the client accepting it • More clear. A single recommendation is too ambiguous Nir Shapira et al, Celeno Communications

  9. Sounding Feedback MCS Control Proposition • MCS recommendation carried in HTC field like in 11n FLA mechanism • Beamformer sends unsolicited MCS recommendations • Use one of the HTC reserved bits to signal MCS recommendation applies only to sounding feedback frames Nir Shapira et al, Celeno Communications

  10. Conclusion • Smart rate selection for MCS of sounding feedback can significantly reduce sounding overhead • The beamformer is the best entity to select and control the MCS • There are obvious advantages for separating the MCS recommendation for traffic and for sounding feedback • We suggest to use the already defined FLA mechanism to additionally recommend MCS for sounding feedback Nir Shapira et al, Celeno Communications

  11. Straw Poll #1 • Do you support having a separate MCS recommendation for uplink traffic and for the explicit sounding feedback frame? Nir Shapira et al, Celeno Communications

  12. Pre Motion #1 • Do you support using one of the reserved bits in HTC to signal that the current Link Adaptation Control field is used for recommending an MCS for the explicit sounding feedback frame, and update the Specification Framework document accordingly? Nir Shapira et al, Celeno Communications

  13. References [1] H. Zhang et al., Explicit Sounding and Feedback, IEEE 802.11-10/1105r0, September. 2010 Nir Shapira et al, Celeno Communications

More Related