May 2011 doc IEEE 802 11 110668 r

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 TX Mask Shoulders vis-à-vis ACI Date: 2011 -05 -04 Authors: Submission Slide 1 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 Abstract • • • A comment requesting a change to the way spectral mask levels are calculated was submitted [1] The comment states that due to the increase in 11 ac in the allowable inband ripple from +/-2 d. B to +/-4 d. B there is a risk that the adjacent channel leakage (which is measured relative to peak power) will also increase by 2 d. B if device manufacturers take advantage of the increased allowable in-band ripple and hence will increase interference and reduce network Tput We therefore study here the effect on system Tput of increased adjacent channel leakage by 2 d. B to quantify the potential degradation We use PHY system simulations to compare the effect of different CCA levels on system Tput as in [2] We also show results of some MAC-level system simulations Submission Slide 2 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 Spectral Mask Submission Slide 3 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 PHY System Simulation Parameters • • • 36 APs and 4 STA per AP are dropped in an area of size 300 x 300 ft and 600 x 600 ft APs are placed regularly with 5 ft std STAs are associated with closest AP according to path loss (which includes random shadowing) One valid transmission per BSS is assumed BSSs are chosen randomly that meet CCA rules (50% probability to choose an AP as transmitter) After all transmitters were chosen SINR is calculated at each receiver and mapped to MCS SISO links over one 40 MHz channel are assumed with 15 d. Bm transmit power CCA level is fixed at -79 d. Bm 4 different frequencies are assigned to the 36 AP in two different ways: – Ordered frequency allocation • 242424 • 131313 • 424242 • 313131 • 242424 • 131313 – Random frequency allocation – in each drop each BSS randomly chooses one of four frequencies Simulation uses 50 drops and 250 TXOP per drop. In each TXOP a maximum number of transmitters across all frequencies are chosen without violating the CCA levels The Baseline ACL was assumed either -25 d. Br or -40 d. Br for all other three frequencies to separately reflect the average interference values with adjacent and alternate adjacent ACL. Tput loss is then measured with -23 d. Br and -38 d. Br Submission Slide 4 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 Results Submission Slide 5 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Parameters Case 5 • Case 5: – Office environment • Approx 30 x 50 m • One floor simulated – 2 BSS • 1 BSS on each of two adjacent channels • AP 0 + 10 clients • AP 11 + 10 clients – TCP flows in both directions for every AP-client pair – ED-CCA level = -62 d. Bm – CRS level = -90 d. Bm Submission Slide 6 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 5 Submission Slide 7 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 5 Submission Slide 8 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC_SIM Results Case 5 • • • Submission s_425_A 23_5 G_B_20_2 x 2_R_e 10_c 62 = s_425_A 25_5 G_B_20_2 x 2_R_e 10_c 62 = s_425_A 38_5 G_B_20_2 x 2_R_e 10_c 62 = s_425_A 40_5 G_B_20_2 x 2_R_e 10_c 62 = s_425_A 99_5 G_B_20_2 x 2_R_e 10_c 62 = Slide 9 82. 29 Mbps 83. 44 Mbps 84. 61 Mbps 84. 67 Mbps 92. 30 Mbps Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Parameters Case 6 • Case 6: – Same floor area as Case 1 – 2 BSS • 1 BSS on one channel, 1 BSS on a second channel (adjacent) – 2 clients per BSS • 6 STA total AP plus clients – TCP flows in both directions for every AP-client pair – ED-CCA level = varied from -53 d. Bm to -71 d. Bm – CRS level = -90 d. Bm Submission Slide 10 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 6 Submission Slide 11 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 6 Submission Slide 12 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC_SIM Results Case 6 Submission Slide 13 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Parameters Case 7 • Case 7: – Same floor area as Case 1 – 2 BSS • 1 BSS on one channel, 1 BSS on a second channel (adjacent) – 2 clients per BSS • 6 STA total AP plus clients • Note client placement – Flows: • • Submission U = 4 x TCP flows only UPLINK for each AP-client pair D = 4 x TCP flows only DOWNLINK for each AP-client pair UD = 2 x TCP flows: AP 0 -> C 2, C 4 -> AP 3 DU = 4 x TCP flows: AP 0 -> C 2, C 1 -> AP 0, C 4 -> AP 3, AP 3 -> C 5 Slide 14 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 7 Submission Slide 15 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 7 Submission Slide 16 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC_SIM Results Case 7 Submission Slide 17 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Parameters Case 8 • Case 8: – Slightly different from case 7 – APs are farther apart • Attempt to get main link margin to be smaller – 2 BSS • 1 BSS on one channel, 1 BSS on a second channel (adjacent) – 2 clients per BSS • 2 AP and 4 clients (effectively only two clients) – 2 TCP flows: • U = C 4 -> AP 3, C 2 -> AP 0 • D = C 4 <- AP 3, C 2 <- AP 0 • UD =C 4 -> AP 3, AP 0 -> C 2 Submission Slide 18 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 8 Submission Slide 19 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 8 Submission Slide 20 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC_SIM Results Case 8 Submission Slide 21 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Parameters Case 9 -C 2 R 4 • Case 9 -C 2 R 4: – 3 Floor office building – 9 BSS per floor (2500 sq ft per BSS maximum) • Semi-rigid AP locations with random variance • Random channel assignment from 2 adjacent channels – 1 -3 clients per BSS (randomly assigned, randomly located) • Over 110% of BSS area – TCP flows: • 3: 1 ratio DOWN to UP, randomly assigned – One flow per client – CCA level = -62 d. Bm – R 4, R 5, R 6 = distinct randomizations Submission Slide 22 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 9 -C 2 R 4 common color = common channel Submission Slide 23 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 9 -C 2 R 4 common color = common channel Submission Slide 24 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 9 -C 2 R 4 Submission Slide 25 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 9 -C 2 R 5 Submission Slide 26 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 9 -C 2 R 5 Submission Slide 27 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 9 -C 2 R 6 Submission Slide 28 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 9 -C 2 R 6 Submission Slide 29 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 9 -C 3 R 7 3 Channels to choose from Submission Slide 30 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 9 -C 3 R 7 Submission Slide 31 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 9 -C 4 R 8 Four channels to choose from Submission Slide 32 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim RXPWR Case 9 -C 4 R 8 Submission Slide 33 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Results Case 9 Submission Slide 34 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Parameters Case 10 • Case 10: – 1 floor – 2 BSS • • • Semi-rigid AP locations with random variance 2 adjacent channels Varying TX Mask shoulders Randomized placements Randomized up and down pair flows – 3: 1 ratio DOWN to UP, randomly assigned – One flow per client • Cases 10. 0 = typical AP separation = 7 m • Cases 10. 1 = typical AP separation = 11 m Submission Slide 35 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Results Case 10. 0 20 randomized cases Submission Slide 36 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Results Case 10. 0 20 randomized cases averaged Submission Slide 37 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Results Case 10. 1 20 randomized cases Submission Slide 38 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Results Case 10. 1 20 randomized cases averaged Submission Slide 39 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 Simulation Conclusions • • • Minimal per-link Tput degradation. No sum network Tput degradation for increase from -40 to -38 d. Br Small (few % points) network Tput degradation for increase from -25 to -23 d. Br probably due to reduced average number of concurrent transmissions • Notes: – – – • In reality when many channels are available (as in the case of 40 MHz channels) the interference level will be mostly -40 d. Br since most channels are not adjacent. Interference level of -25 d. Br will only be the dominant case if only 2 -3 channels are available The simulation assumes full buffer – all nodes always have something to transmit – this is a worst case scenario from interference point of view. While it’s not clear that implementations of 802. 11 ac will actually have in-band signal power variations of +/-4 d. B, the effect on network Tput is modest. Therefore, we propose not to include restrictions (equations, etc. ) to the spectral flatness. Submission Slide 40 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 References [1] 11 -11 -0276 -11 -00 ac-tgac-d 0 -1 -comments. xls – CID 488 [2] 11 -11 -0061 -00 -00 ac-cca-threshold-levels. ppt Submission Slide 41 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 Appendix • Randomly selected topographies Submission Slide 42 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 10. 0 -0 Submission Slide 43 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 10. 0 -16 Submission Slide 44 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 10. 0 -17 Submission Slide 45 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 10. 1 -0 Submission Slide 46 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 10. 1 -16 Submission Slide 47 Matt Fischer, Broadcom

May 2011 doc. : IEEE 802. 11 -11/0668 r 2 MAC Sim Topography Case 10. 1 -17 Submission Slide 48 Matt Fischer, Broadcom