00:00

Benefit Analysis of Unequal Modulation in WLAN Channels

This document presents an analysis of the benefits of Unequal Modulation (UEQM) in WLAN channels, focusing on improving Bit Error Rate (BER) without Beamforming (BF), and exploring performance enhancements in Link Adaptation. It discusses the effect of UEQM on BER, analysis setup, simulation results, and implications for MCS selection and rate adaptation strategies in wireless communication systems.

aubia
Download Presentation

Benefit Analysis of Unequal Modulation in WLAN Channels

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. March 2024 doc.: IEEE 802.11-24/0438r0 UEQM Benefit Analysis Date: 4-Mar-24 Authors: Name Sigurd Schelstraete Affiliations MaxLinear email sschelstraete@maxlinear.com Rainer Strobel rstrobel@maxlinear.com Submission Slide 1 Sigurd Schelstraete, MaxLinear

  2. March 2024 doc.: IEEE 802.11-24/0438r0 Abstract Previous submissions: • [1] discusses the channel characteristics of MIMO WLAN channels, reviews 11n unequal modulation and shows ~2dB gain for UEQM on the 2x2 D NLOS channel over EQM with power scaling (enforcing 2 streams) • [2] discusses unequal MCS vs. UEQM and shows UEQM gain of ~2dB on the 2x2 D NLOS channel (enforcing 2 streams) • [3] presents performance gains on the 2x2, 4x2 and 4x4 D NLOS and B LOS channel, showing 1-5dB gain (enforcing 2 or 4 streams, respectively) • [4] compares UEQM with all possible modulation/code rate combinations with UEQM limited to existing combinations and analyses the 4x2 D NLOS case In this submission: • Understanding the underlying benefit of UEQM • Understanding performance of UEQM without BF • Possible enhancements to Link Adaptation • • Submission Slide 2 Sigurd Schelstraete, MaxLinear

  3. March 2024 doc.: IEEE 802.11-24/0438r0 Effect of UEQM on BER • The ultimate goal is improving BER (or PER, …) at the decoder output • Without BF: • Streams have similar SNR • UEQM could lower bit errors at decoder input for one of the streams • How much of an effect? • With BF: • Streams have different SNR • Modulation could be “matched” to SNR • No need to equalize BERs of different streams at decoder input for optimal performance Submission Slide 3 Sigurd Schelstraete, MaxLinear

  4. March 2024 doc.: IEEE 802.11-24/0438r0 Analysis setup • Studying UEQM on “controlled” channels • Initial simulations on Identity channel • With tunable difference in SNR between the streams • Allows for controlled behavior of the channel to understand UEQM • Behavior is then verified on actual (model B) channels with similar SNR difference between streams Submission Slide 4 Sigurd Schelstraete, MaxLinear

  5. March 2024 doc.: IEEE 802.11-24/0438r0 (1) UEQM without BF • MCS 4 (16 QAM, Rate ¾), MCS 2 (QPSK, Rate ¾) vs. unequal modulation (16 QAM + QPSK, Rate ¾) tested • Effective rates are 6 bit/carrier, 3 bit/carrier and 4.5bit/carrier respectively • No BF  No SNR difference between spatial streams • UEQM shows only a small BER improvement • at the cost of 25% lower rate • BER improvement, but no benefit in throughput Submission Slide 5 Sigurd Schelstraete, MaxLinear

  6. March 2024 doc.: IEEE 802.11-24/0438r0 (2) UEQM for varying SNR difference SNR difference of 6 and 12 dB between streams • Use BF for transmission With increasing SNR difference, BER of UEQM improves relative to EQM • At 12 dB difference 16 QAM + QPSK outperforms QPSK + QPSK ! • • Submission Slide 6 Sigurd Schelstraete, MaxLinear

  7. March 2024 doc.: IEEE 802.11-24/0438r0 (2) UEQM for varying SNR difference – cont. • Required SNR for the various modulations as funtion of SNR difference • At 12 dB difference 16 QAM + QPSK outperforms QPSK + QPSK • Detailed analysis on next slides Submission Slide 7 Sigurd Schelstraete, MaxLinear

  8. March 2024 doc.: IEEE 802.11-24/0438r0 BER Analysis AWGN Channel 6dB and 12dB SNR difference Combined BER is dominated by the spatial stream with higher BER With a large SNR difference (e.g. 12dB), increasing the constellation size (QPSK  16 QAM) on the good spatial stream improves the overall BER • more error-free bits on the good stream • same errors on the bad stream Conclusion: SNR difference between streams is an important driver for MCS selection, but decoder effects need to be considered as well • Implications for rate adaptation • • • Submission Slide 8 Sigurd Schelstraete, MaxLinear

  9. March 2024 doc.: IEEE 802.11-24/0438r0 WLAN Channel Example without BF Simulation Results • Without beamforming, the SNR difference between spatial steams is small • The same behavior as for the AWGN channel is observed: • BER can be lowered on one of the streams, but at significant performance penalty • The required SNR is lowered by less then 1 dB from mcs 4 (6 bit/s/Hz) to mcs 4/2 (4.5 bit/s/Hz) • There is no performance (throughput) advantage advantage unequal modulation Submission Slide 9 Sigurd Schelstraete, MaxLinear

  10. March 2024 doc.: IEEE 802.11-24/0438r0 WLAN Channel Example: B LOS 2x2 Simulation Results • For 2x2 B LOS channel with beamforming, the SNR difference between spatial streams is around 12dB • The same behavior as for the AWGN channel is observed: Unequal modulation MCS 4/2 achieves a higher rate at a lower SNR than 2SS MCS2 Submission Slide 10 Sigurd Schelstraete, MaxLinear

  11. March 2024 doc.: IEEE 802.11-24/0438r0 Implications for Link Adaptation • The previous results allow us to identify possible enhancements in Link Adaptation • Better exploit the understanding of UEQM performance • See next three slides for proposal and comparison with existing feedback mechanisms Submission Slide 11 Sigurd Schelstraete, MaxLinear

  12. March 2024 doc.: IEEE 802.11-24/0438r0 Link Adaptation Option I – Using existing feedback • Available: PER from acknowledgements • Test all relevant constellation size/code rate settings for equal and unequal modulation • Advantage: • No additional measurements or evaluation required • Disadvantages: • Slow, because many options must be tested Submission Slide 12 Sigurd Schelstraete, MaxLinear

  13. March 2024 doc.: IEEE 802.11-24/0438r0 Link Adaptation Option II – Using existing feedback • Available: SNR per spatial stream from sounding, PER from acknowledgements • Use SNR difference from sounding for initial MCS pick and adjust based on PER • Advantage: • Minimal difference to existing link adaptation • Disadvantages: • Optimal setting is not necessarily tested • SNR difference from sounding may not be accurate Submission Slide 13 Sigurd Schelstraete, MaxLinear

  14. March 2024 doc.: IEEE 802.11-24/0438r0 Link Adaptation Option III – Additional feedback Available: PER from acknowledgements, SNR per spatial stream from sounding, SNR margin per spatial stream (each data packet, not available from sounding) Pick MCS based on sounding SNR and SNR gap per spatial stream Update SNR gap, based on PER and SNR margin feedback Advantages: • Fast adaptation • Accurate decision, because of per-spatial stream feedback Disadvantages: • New feedback from the receiver required • • • • • Submission Slide 14 Sigurd Schelstraete, MaxLinear

  15. March 2024 doc.: IEEE 802.11-24/0438r0 Link Adaptation Comparison Unequal modulation with the modulation difference derived from sounding (Option II) achieves higher rates than equal modulation, but results in steps in the rate-reach curve (like equal modulation). With SNR margin feedback (Option III), higher rates and a more smooth rate-vs.-reach curve is achieved. The performance is equal to a full search, but less time consuming. • • Submission Slide 15 Sigurd Schelstraete, MaxLinear

  16. March 2024 doc.: IEEE 802.11-24/0438r0 Conclusions • We investigated how UEQM contributes to improve BER in various situations • SNR difference between streams provides guidance on choice of UEQM, but feedback can be refined • Rate adaptation may benefit form (near-) instantaneous feedback of SNR margin Submission Slide 16 Sigurd Schelstraete, MaxLinear

  17. March 2024 doc.: IEEE 802.11-24/0438r0 References [1] UHR MIMO RvR enhancement with unequal modulation, 24/0016 [2] Unequal Modulation in MIMO TxBF in 11bn, 24/0113 [3] Improved Tx Beamforming with UEQM, 24/0117 [4] Unequal Modulation over Spatial Streams, 24/0176 Submission Slide 17 Sigurd Schelstraete, MaxLinear

More Related