1 / 20

System Level Simulations on Beamforming and Link Adaptation

Date: 2014-05-11. System Level Simulations on Beamforming and Link Adaptation. Authors:. Transmit beamforming (TxBF) improves the link robustness. A significant performance improvements achieved by TxBF was shown in [1] in an artificial OBSS scenario, by actual over-the-air measurements.

mio
Download Presentation

System Level Simulations on Beamforming and Link Adaptation

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. Date: 2014-05-11 System Level Simulations on Beamforming and Link Adaptation Authors: Yan Zhang (Marvell)

  2. Transmit beamforming (TxBF) improves the link robustness. • A significant performance improvements achieved by TxBF was shown in [1] in an artificial OBSS scenario, by actual over-the-air measurements. • This presentation examines the TxBF gain in more realistic OBSS scenarios using SLS. • Link adaptation is essential to the system level performance. • Ideal link adaptation provides an upper bound on system level performance. • ACK-based link adaptation provides a practical system level performance. • How much gap exists between performance achieved by different link adaptations? • Integrated PHY/MAC system level simulations investigate the system level gain achieved by TxBF and the impact of link adaptation. Overviews Yan Zhang (Marvell)

  3. Use integrated MAC + PHY simulation • Simulation Time: 5 seconds • Downlink traffic only • Modeling: application, MAC and PHY layers • Application layer • Each STA runs a UDP server and the associated AP runs the corresponding UDP client • UDP CBR traffic: 1.2 Mbps per STA for downlink traffic • Best Effort: AC = 2 in EDCA implemented in MAC • IPv4 stack • Static ARP: no ARP message exchange (some management frames could be easily lost in dense 802.11 network) Integrated SLS Setup (1) Yan Zhang (Marvell)

  4. MAC layer • TBTT: uniform random distributed none of the BSS is synchronized • Rate Adaptation: • Ideal link adaptation • Minstrel algorithm [2] • PHY layer • RBIR PHY abstraction [3] • 20MHz channel bandwidth Integrated SLS Setup (2) Yan Zhang (Marvell)

  5. A genie receiver knows: • Who are the active interferers (and the portion of frame overlapped among all interferences) • The effective channels from the desired and interfering transmitters. • Beamforming vector is selected to maximizes SNR (not SINR). • Only based on channel linked to desired receiver • The MCS is selected to maximize goodput. • PER per MCS is calculated according to the effective channels from the desired and interfering transmitters. Ideal Link Adaptation Yan Zhang (Marvell)

  6. Small indoor hotspot using the PHY and MAC parameters/requirements defined in [4] • 19 BSSs and 10 STAs per BSS with reuse factor 3 • 4x1 channels • Channel B for STA to STA link. • Channel D for AP to AP, AP to STA link Scenario 3 Assumptions Yan Zhang (Marvell)

  7. Scenario 3: Average BSS Throughput 19% 35% Note that the gain is TxBF over No n-TxBF for ideal and minstrel LA respectively Yan Zhang (Marvell)

  8. Scenario 3: Typical STA Throughput Note that the gain is TxBF over Non-TxBF or ideal and minstrel LA respectively Yan Zhang (Marvell)

  9. Scenario 3: STA Throughput Distribution Yan Zhang (Marvell)

  10. Scenario 3: Tone SINR statistics 10dB Yan Zhang (Marvell)

  11. Large outdoor hotspot using the PHY and MAC parameters/requirements defined in [4] • 19 BSSs and 10 STAs per BSS with reuse factor 1 • 4x1 channels • ITU UMI channel for all links • 100% outdoor STAs • Only ideal link adaptation is simulated. Scenario 4 Assumption Yan Zhang (Marvell)

  12. Scenario 4: Average BSS Throughput Yan Zhang (Marvell)

  13. Scenario 4: STA Throughput Distribution Note that about 6% of STAs have zero throughput during the entire simulation due to either extremely low SINR or no chance of transmission. Yan Zhang (Marvell)

  14. Scenario 4: Typical STA Throughput Yan Zhang (Marvell)

  15. Scenario 4: Tong SINR Distribution 12.5dB Yan Zhang (Marvell)

  16. Scenario 4: Wraparound vs. Non-Wraparound • Average BSS throughput of different tiers have similar statistics with wraparound. • Without wraparound, outer tier BSSs have much higher throughput than the inner tier BSSs due to much smaller OBSS interference. Yan Zhang (Marvell)

  17. Scenario 4: Wraparound vs. Non-Wraparound (2) • Non-associated STAs: 3% (wrap-around) vs. 6% (without wrap-around). • More STAs has larger throughput without wrap-around, since STAs in the second and third tier experience smaller interference. Yan Zhang (Marvell)

  18. TxBF significantly improves throughput, especially 5%-tile STA throughput. • The TxBF gain of average BSS throughput is about 19% with ideal link adaptation, and about 35% with minstrel link adaptation for small indoor scenario. • The TxBF gain of 5% tile STA throughput is about 22% with ideal link adaptation, and about 39% with minstrel link adaptation for small indoor scenario. • The TxBF gain of average BSS throughput is about 34%, and gain of 10% tile STA throughput is about 103% for large outdoor scenario. • 11ac single user TxBF should be adopted as the baseline system level simulation evaluation for 11ax. Conclusions and Future study Yan Zhang (Marvell)

  19. BSS throughput employing ideal link adaptation is about 4 times higher than that of employing ACK-based link adaptation. • Which type of link adaptation should be used for system level simulation calibration? • There are a small amount of STAs are not able to associate with APs in current large outdoor scenario. • Wrap-around can be adopted to improve association / provide balanced interference and throughput. Conclusions and Future study Yan Zhang (Marvell)

  20. [1] 11-13-1126-00-0hew-beamforming-under-OBSS-interference [2]http://wireless.kernel.org/en/developers/Documentation/mac80211/RateControl/minstrel [3] 11-14-0117-00-0hew-phy-abstraction-for-hew-system-level-simulation [4] 11-13-1001-06-0hew-simulation-scenarios-document-template Reference Yan Zhang (Marvell)

More Related