Doc.: IEEE 802.11-14/1207r0 Submission Imad Jamil (Orange)Slide 1 OBSS reuse mechanism which preserves fairness Date: 2014-09-15 Authors: September 2014.

Slides:



Advertisements
Similar presentations
Submission doc.: IEEE 11-14/0868r0 July 2014 Johan Söder, Ericsson ABSlide 1 UL & DL DSC and TPC MAC simulations Date: Authors:
Advertisements

Dynamic Sensitivity Control V2
Submission doc.: IEEE /1225r1 Considerations on CCA for OBSS Opearation in ax Date: Slide 1Huawei Authors:
Doc.: IEEE /0025r0 Submission Jan 2015 Dynamic Sensitivity Control Roaming Date: 2015-January Authors: Graham Smith, SR TechnologiesSlide 1.
Discussion on The Receiver Behavior for DSC/CCAC with BSS Color
Submission doc.: IEEE /0374r0 Mar 2015 John Son, WILUS InstituteSlide 1 Further Considerations on Legacy Fairness with Enhanced CCA Date:
Submission doc.: IEEE 11-15/613r0 May 2015 Chinghwa Yu et al, MediaTek Inc.Slide 1 Box 5 Calibration Result Date: Authors:
Doc.: IEEE Amin Jafarian, Newracom 1 CCA Regime Evaluation Revisited March 2015 NameAffiliationsAddressPhone Amin
Doc.: IEEE /0319r1 Submission March 2015 Takeshi Itagaki, Sony CorporationSlide 1 Impact of TPC coupled to DSC for legacy unfairness issue Date:
Doc.: IEEE /0861r0 SubmissionSayantan Choudhury Impact of CCA adaptation on spatial reuse in dense residential scenario Date: Authors:
Doc.: IEEE /1207r1 Submission Imad Jamil (Orange)Slide 1 OBSS reuse mechanism which preserves fairness Date: Authors: September 2014.
Doc.: IEEE r1 Amin Jafarian, Newracom 1 CCA Regime Evaluation Revisited March 2015 NameAffiliationsAddressPhone Amin
Submission doc.: IEEE /0085r1 Jan 2015 John Son, WILUS InstituteSlide 1 Legacy Fairness Issues of Enhanced CCA Date: Authors:
Doc.: IEEE /1443r0 SubmissionEsa Tuomaala Adapting CCA and Receiver Sensitivity Date: Authors: Slide 1 November 2014.
Doc.: IEEE /1420r1Nov 2014 Submission Po-Kai Huang (Intel) Slide 1 The Impact of Preamble Error on MAC System Performance Date: NameAffiliationsAddressPhone .
Doc.: IEEE /1227r3 SubmissionSlide 1 OFDMA Performance Analysis Date: Authors: Tianyu Wu etc. MediaTek Sept 2014 NameAffiliationsAddressPhone .
Discussion on OFDMA in IEEE ax
802.11ax scenario 1 CCA Date: Authors: March 2015
Submission doc.: IEEE /0623r1 May 2015 Guido R. Hiertz, Ericsson et al.Slide 1 TGax simulation scenario “Box 5” – calibration results Date:
Doc.: IEEE /0116r0 SubmissionYakun Sun, et. Al.Slide 1 Long-Term SINR Calibration for System Simulation Date: Authors: NameAffiliationsAddressPhone .
Doc. No. IEEE hew-r1 Submission July 2013 Klaus Doppler, NokiaSlide 1 Evaluation Criteria and Simulation Scenarios Date: July 16, 2013 Authors:
Submission doc.: IEEE /0374r1 Mar 2015 John Son, WILUS InstituteSlide 1 Further Considerations on Legacy Fairness with Enhanced CCA Date:
Doc.: IEEE /1153r0 Submission September 2013 Laurent Cariou (Orange)Slide 1 Simulation scenario proposal Date: Authors:
Doc.: IEEE /1081r0 SubmissionSayantan Choudhury HEW Simulation Methodology Date: Sep 16, 2013 Authors: Slide 1.
Doc.: ax Submission Sept 2014 Slide 1 Effect of CCA in residential scenario part 2 Date: Authors:
Doc.: IEEE /0804r0 Submission July 2015 TG ax Outdoor Enterprise Scenario and DSC Date: Authors: Graham Smith, SR TechnologiesSlide 1.
Doc.: ax Submission July 2014 Slide 1 Proposed Calibration For MAC simulator Date: Authors:
Doc.: IEEE /0637r0 Submission May 2014 James Wang et. al., MediaTekSlide 1 Spatial Reuse and Coexistence with Legacy Devices Date:
Doc.: IEEE /0523r0 Submission April 2014 Imad Jamil (Orange)Slide 1 MAC simulation results for Dynamic sensitivity control (DSC - CCA adaptation)
Doc.: IEEE r0 Amin Jafarian, Newracom 1 CCA Revisit May 2015 NameAffiliationsAddressPhone Amin
Doc.: IEEE /610r1 Submission Vida Ferdowsi, Newracom May 2015 Slide 1 NameAffiliationsAddressPhone Vida Ferdowsi Daewon Lee Reza Hedayat.
Doc.: IEEE /0799r2 Submission June 2014 Nihar Jindal, Broadcom Modifications to Simulation Scenarios and Calibration Process Date:
Doc.: IEEE /0680r1 SubmissionJiyong Pang, Huawei TechnologiesSlide 1 Reference Box5 Calibration Assumptions and Parameters Date: Authors:
Doc.: IEEE / ax Submission Eduard Garcia-Villegas Drivers of the dynamic CCA adaptation Authors: Nov Date:
Doc.: IEEE /0779r2 Submission June 2014 Dynamic Sensitivity Control Practical Usage Date: 2014-July Authors: Graham Smith, DSP GroupSlide 1.
Doc.: IEEE /0542r0 SubmissionSimone Merlin, QualcommSlide 1 HEW Scenarios and Goals Date: Authors: May 2013.
Submission doc.: IEEE 11-13/1401r0 Nov Josiam, Kuo, Taori et.al., SamsungSlide 1 System Level Assessments for Outdoor HEW Deployments Date: YYYY-MM-DD.
Doc.: IEEE / ax Submission M. Shahwaiz Afaqui DSC calibration results with NS-3 Authors: Nov
Doc.: IEEE / ax Submission M. Shahwaiz Afaqui DSC calibration results with NS-3 Authors: Nov
Submission doc.: IEEE /1373r1 November 2015 Narendar Madhavan, ToshibaSlide 1 Updated Box 5 Calibration Results Date: Authors:
Doc.: IEEE /0212r3 Submission Feb 2016 TG ax Enterprise Scenario, Color and DSC Date: Authors: Graham Smith, SR TechnologiesSlide 1.
Doc.: IEEE /1392r5 SubmissionSuhwook Kim, LG ElectronicsSlide 1 Simulation results for Box 5 calibration Date: Authors: NameAffiliationsAddressPhone .
Doc.: IEEE /0635r1 Submission May 2014 Dynamic Sensitivity Control Implementation Date: 2014-May Authors: Graham Smith, DSP GroupSlide 1.
Doc.: IEEE /XXXr0 Submission May 2004 Masahiro Takagi and Tomoko Adachi, TOSHIBASlide 1 Simulation Scenarios and Comparison Criteria for Coexistence.
SubmissionJoe Kwak, InterDigital1 PHY measurements for interference reduction from 11h Joe Kwak, Marian Rudolf InterDigital doc: IEEE /537r0July.
Submission doc.: IEEE /0871r1 Jul Jiyong Pang, et. al. Huawei Further Calibration Results towards Integrated System Level Simulation Date:
OFDMA performance in 11ax
Month Year doc: IEEE /xxxxr0
OFDMA performance in 11ax
Simulation results for spatial reuse in 11ax
11ax PAR Verification using UL MU-MIMO
Simulation-based evaluation of DSC in enterprise scenario
TG ax A Unified Approach to Spatial Reuse
TG ax Indoor Enterprise Scenarios, Color, DSC and TPC
Performance Evaluation of OBSS Densification
OFDMA performance in 11ax
2840 Junction Ave. San Jose, CA 95134
Simulation Analysis of ED Threshold Levels
Joint submission for Box 5 calibration
TG ax A Unified Approach to Spatial Reuse
Considerations on CCA for OBSS Opearation in ax
802.11ax scenario 1 CCA Date: Authors: March 2015
Box 5 Calibration Result
Month Year doc.: IEEE /0578r0 May 2016
802.11ax scenario 1 CCA Date: Authors: March 2015
TG ax A Unified Approach to Spatial Reuse
System Level Simulator Evaluation with/without Capture Effect
AP Coordination in EHT Date: Authors: Name Affiliations
DSC Calibration Result
802.11ax scenario 1 CCA Date: Authors: March 2015
Presentation transcript:

doc.: IEEE /1207r0 Submission Imad Jamil (Orange)Slide 1 OBSS reuse mechanism which preserves fairness Date: Authors: September 2014

doc.: IEEE /1207r0 Submission Context It has been shown in many presentations that CCA adaptation (DSC, fixed CCA) can lead to very strong per-user throughput increase in dense environments (multiple x-times improvements) We look at it for a planned freq reuse 3 deployment (close to scenario 3), In 523r0, we demonstrated the benefits of DSC, but show some fairness warning (even for 11ax devices) In this presentation –We present fairness analysis with different CCA adaptation (DSC, fixed CCA), with TPC –We propose a mode that optimizes both area throughput and fairness Imad Jamil (Orange)Slide 2 September 2014

doc.: IEEE /1207r0 Submission DSC algorithm DSC (CCA control): –each STA adjust its CCA to CCA = Useful Rx Power – Margin –AP adjust its CCA to CCA = Useful Rx Power from further STA - Margin CCA adapation is caped by max CCA value (upper limit) Imad Jamil (Orange)Slide 3 STAs at 3m: CCA=-50dBm STAs at 5m: CCA =-60dBm APs: CCA =-60dBm margin 20dB Max CCA -50dBm STAs at 3m: CCA=-40dBm STAs at 5m: CCA =-60dBm APs: CCA =-60dBm margin 20dB Max CCA -40dBm September 2014

doc.: IEEE /1207r0 Submission TPC algorithm TPC (Transmit Power control): –each receiving STA requests transmitting STA to adjust its TP so that Rx Power = CCA + Margin We took a margin of 20dB Imad Jamil (Orange)Slide 4 September 2014

doc.: IEEE /1207r0 Submission Tx :15 dBm [21 m] Close to scenario 3 - First tier only (7 BSSs), 8 STAs per BSS, frequency reuse 3 pattern Simulation scenario Imad Jamil (Orange)Slide 5 [7 m] September 2014

doc.: IEEE /1207r0 Submission Parameters PHY o IEEE n o Path Loss: ITU UMi ( log10(d)+ 21log10(2400/900MHz) ) o Band: 5 Ghz o Channel: 20 MHz o Tx power: 15 dBm Traffic o Full buffer UDP traffic o DL (AP->STA): 4Mbps per STA o UL (STA->AP): 4Mbps per STA Metric Aggregate throughput, per user throughput CDF Rate adaptation AARF with selection of set of MCSs (link adaptation) Simulation scenario Imad Jamil (Orange)Slide 6 September 2014

doc.: IEEE /1207r0 Submission How to improve fairness (1/2) with CCA adaptation DSC presents unfairness issues at saturation state, linked to the fact that different users have different CCA parameters How to improve fairness? –1: DSC unfairness can be improved by tuning the max CCA threshold –2: CCA can also be fixed for the entire zone We propose to compare these approaches and evaluate fairness –between 11ax STAs –with regards to legacy STAs Slide 7 September 2014

doc.: IEEE /1207r0 Submission 1 - DSC fairness improvements with max CCA threshold optimization Slide 8 Unfairness between STAs at different range in UL We changed the max CCA threshold to limit the CCA difference between STAs, which limits the unfairness STAs at 3m: CCA=-50dBm STAs at 5m: CCA =-60dBm APs: CCA =-60dBm margin 20dB Max CCA -50dBm STAs at 3m: CCA=-40dBm STAs at 5m: CCA =-60dBm APs: CCA =-60dBm margin 20dB Max CCA -40dBm September 2014

doc.: IEEE /1207r0 Submission 2 - Fairness comparison with Fixed CCA Slide 9 Same simulation scenario (7 11ax, 1 legacy per AP), comparing: –fixed CCA: -82, -70, -60, -50dBm –DSC with max CCA threshold (-50dBm): (STAs at 3m: CCA=-50dBm, STAs at 5m: CCA =-60dBm, APs: CCA =-60dBm) 11ax STAslegacy STAs : best fixed CCA -50 CCAC: DSC UL User throughput CDF results for 11ax devices: Curve slope is a good indication of fairness September 2014

doc.: IEEE /1207r0 Submission 2 - Fairness comparison with Fixed CCA Slide 10 Conclusion: at saturation state: –fixed CCA is not more fair than DSC between 11ax STAs –fixed CCA is unfair with legacy devices Without legacy STAsWith legacy STAs September 2014

doc.: IEEE /1207r0 Submission How to improve fairness (2/2) joint CCA/TPC adaptation DSC and TPC both present unfairness issues, linked to the fact that different users have different CCA/TP parameters –CCA control favors 11ax, TP control favors legacy STAs How to improve fairness? –propose a scheme that adapts both CCA and TP, following a specific rule in order that each mechanism (CCAC and TPC) counteracts the unfairness from the other Slide 11 September 2014

doc.: IEEE /1207r0 Submission Reuse scheme: balanced CCAC/TPC Example with an equal modification of CCA and TP Slide 12 STA calculates a Delta_x value, which is the amount of dB by which CCA and TP will be changed –this delta_x value would be equal to delta_CCA if CCA adaptation only is performed and to delta_TP if TPC only is performed –if the STA uses the concept of DSC on this first Delta_x calculation, we have: Delta_x = Rx_power – Margin – standard_CCA (ex: Margin= 20dB, standard_CCA = -79dBm for 40MHz) STA applies a specific formula for the calculation of delta_CCA and delta_TP, based on Delta_x –in our example, we considered the rule as being an equal change of CCA and TP Delta_CCA=Delta_TP=Delta_x / 2 STA applies CCA adaptation using Delta_CCA STA requests AP to apply TP adaptation using Delta_TP –or applies TP adaptation using Delta_TP September 2014

doc.: IEEE /1207r0 Submission Fairness comparison with balanced CCAC/TPC (1/4) Slide 13 Same simulation scenario, comparing: –best fixed CCA: -60 dBm –DSC with max CCA threshold (-50dBm): (STAs at 3m: CCA=-50dBm, STAs at 5m: CCA =-60dBm, APs: CCA =-60dBm) –TPC –balanced CCAC/TPC (equal change between CCA and TP) Aggregate throughput results, with and without legacy devies: All solutions present similar results (expect TPC, which collapses in the presence of legacy devices TPC in presence of legacy devices September 2014

doc.: IEEE /1207r0 Submission Slide 14 Same simulation scenario, comparing: –best fixed CCA: -60 dBm –DSC with max CCA threshold (-50dBm): (STAs at 3m: CCA=-50dBm, STAs at 5m: CCA =-60dBm, APs: CCA =-60dBm) –TPC –balanced CCAC/TPC (equal change between CCA and TP) Without legacy STAs (8 11ax STA per AP) With legacy STAs (7 11ax STA and 1 legacy STA per AP Balanced CCAC/TPC and TPC DSC and fixed CCA Balanced CCAC/TPC DSC and fixed CCA UL User throughput CDF results for 11ax devices: Curve slope is a good indication of fairness Fairness comparison with balanced CCAC/TPC (2/4) September 2014

doc.: IEEE /1207r0 Submission Slide 15 Same simulation scenario, comparing: –best fixed CCA: -60 dBm –DSC with max CCA threshold (-50dBm): (STAs at 3m: CCA=-50dBm, STAs at 5m: CCA =-60dBm, APs: CCA =-60dBm) –balanced CCAC/TPC (equal change between CCA and TP) legacy STAs throughput (7 11ax STA and 1 legacy STA per AP) Balanced CCAC/TPC DSC Legacy throughput are better preserved than with DSC UL User throughput CDF results for legacy devices: Fairness comparison with balanced CCAC/TPC (3/4) September 2014

doc.: IEEE /1207r0 Submission Slide 16 Conclusion: –Balanced CCAC/TPC seems the best solution with regards to fairness between 11ax and with legacy devices –It gives equal area throughput gains than CCAC (DSC, fixed CCA..) and TPC Further advantages: –It preserves the benefit of being directly beneficial for the STA to implement (compared to TPC) –It can be very simple to define in 11ax: we can define in 11ax only the proportion rule to respect between Delta_CCA and Delta_TPC: gives every STA the possibility to use its own algorithm (need to see what proportion rule is the optimum) this proportion rule could be parametrized in case of a managed network (need simulations to evaluate if different proportion rules are required depending on the environments, or depending on the proportion of legacy/11ax devices) Note that such proportion rule would need to be adapted when combined with UL OFDMA and UL MU-MIMO, as TP adaptation will be required also for those schemes. Fairness comparison with balanced CCAC/TPC (4/4) September 2014

doc.: IEEE /1207r0 Submission Conclusion Fairness has to be considered for OBSS reuse mechanisms –even if it only arises at saturation state Fairness analysis with CCA control –We previously showed fairness issues with DSC –fixed CCA is not more fair than DSC between 11ax STAs –fixed CCA is unfair with legacy devices New solution to consider: balanced CCAC/TPC –in our simulation scenario, Balanced CCAC/TPC performs way better with regards to fairness between 11ax and with legacy devices –and It gives equal area throughput gains than CCAC (DSC, fixed CCA..) and TPC –Need to evaluate it on other scenarios, and to optimize the ratio between CCAC and TPC Imad Jamil (Orange)Slide 17 September 2014