3 GPP TSGRAN WG 486 b Meeting Melbourne

3 GPP TSG-RAN WG 4#86 b Meeting Melbourne, Australia, Apr 16 – 20, 2018 Agenda: 7. 4. 8. 3. 2 R 4 -180 xxxx Summary of network performance analysis for spherical coverage Samsung

Evaluation assumptions • Assumptions as agreed in R 4 -1801198 & R 4 -1803275. Differences in reference and evaluated spherical coverages are shown below UL peak EIRP (d. Bm) Reference spherical coverage (Ref) Evaluated spherical coverage (Eval) UE orientation Apple/Intel (R 4 -1802220) Qualcomm (R 4 -1805669) Samsung (R 4 -1803820) 22 22. 5 22. 4 TR TR CFG TR (finite beam points) -5 d. B@50%tile -5. 6 d. B@50%tile -5 d. B@50%tile -10 d. B@20%tile -11. 7 d. B@20%tile -10 d. B@20%tile “Model 2 (A+B)” (2 -panel, same “CFG 1” (1 panel) Assumption 1 side) -11. 7 d. B@50%tile -8. 6 d. B@50%tile -11. 5 d. B@50%tile -18. 4@20%tile -15 d. B@20%tile -17. 7 d. B@20%tile Uniform on sphere surface EL: U(0: 180); Az: U(0: 360) Qualcomm Samsung Media. Tek (R 4 -1804412) 22 TR CDF model -5 d. B@50%tile -10 d. B@20%-tile LGE (R 4 -1804489) 21. 5 TR CDF model -5 d. B@50%tile -10 d. B@20%-tile Assumption 3 -11. 6 d. B@50%tile -13. 5 d. B@20%tile -12. 5 d. B@50%tile -18. 5 d. B@20%tile LGE

Summary of outage evaluation results • Outage evaluation results submitted to RAN 4#86 b Outage (UL 20 MHz) Apple/Intel (R 4 -1805381) LGE (R 4 -1804489) Samsung (R 4 -1803820) Media. Tek (R 4 -1804412) Peak EIRP UMa UMa (0% indoor) (20% indoor) (100% indoor) Dense urban ISD ISD ISD (0% indoor) (20% indoor) (100% indoor) 200 m 400 m 22 d. Bm 0% 4% 21. 2 d. Bm 23. 2 d. Bm 21. 2 d. Bm 22. 4 d. Bm 23. 2 d. Bm 0. 3% 0. 2% 0. 1% 0. 0% 2. 5% 1. 6% 1. 3% 1. 0% 0. 9% 22 d. Bm 0% 3. 2% 6% 15% 39% 63% 7. 8% 14. 8% 37. 4% 66. 1% 7. 2% 13. 8% 35. 0% 63. 2% 6. 7% 13. 3% 32. 6% 61. 1% 0% 0. 1% 0. 06% 0% 0% 0% 5% 25% In. H Remark 0% TR + mask 0. 4% 0% 0% UE assumption 1 Specific CDF model

Summary of throughput evaluation results - 1 • Mean throughput evaluation results submitted to RAN 4#86: loss compared to respective reference spherical coverage • No hand/body losses or other blockages is considered in the analysis. It is not necessary if evaluation assumes at least one antenna array is not blocked Apple/Intel (R 4 -1805381) Qualcomm (R 4 -1804660) Samsung (R 4 -1803820) Media. Tek (R 4 -1804412) LGE (R 4 -1804489) Urban macro (0% indoor) ISD 200 m ISD 400 m Dense urban (0% indoor) DL <0. 5% 3% <1% UL 13% 12% DL 4. 6% 9. 9% 7. 1% 11. 5% UL DL 18. 5% 6. 62% 20. 3% 11. 05% 9. 6% 6. 10% 7. 9% 7. 28% UL 5. 51% 7. 58% 7. 61% 5. 10% UL 7. 3% 11. 2% UL 5. 7% 13. 7% Mean throughput loss Indoor hotspots 2. 9% 1. 4% 3. 4%

Summary of throughput evaluation results - 2 • 5%-tile throughput evaluation results submitted to RAN 4#86: loss compared to reference spherical coverage 5%-tile throughput loss Qualcomm (R 4 -1804660) Samsung (R 4 -1803820) DL UL Urban macro (0% indoor) ISD 200 m ISD 400 m 16. 9% 30. 1% 56. 8% N. A. 27. 83% 19. 87% 15. 53% 3. 90% Dense urban (0% indoor) 23. 1% 33. 5% 28. 00% 23. 26% Indoor hotspots 26. 4% 30% 21. 74% 20. 05% • Note: The evaluation is for full buffer traffic and round-robin scheduler.

Network Performance Sensitivity to EIRP CDF Source Apple/Intel (R 4 -1805381) Qualcomm (R 4 -1805669) Deviation from a ref 50%-tile 20%-tile Samsung (R 4 -1803820 rev) Apple/Intel (R 4 -1805381) -8 d. B -5 d. B UL throughput loss wrt TR model Mean 5%-tile > 25% 10% 18% 60% 5% 26% 1. 49%, 0 d. B 5. 51% +5 d. B 8. 74% Qualcomm (R 4 -1805669) Remark Ref for deviation is TR model Ref for deviation is Assumption 1 Samsung (R 4 -1803820 rev)

Appendix: Assumptions for EIRP CDF • Updated assumptions has been agreed in RAN 4 AH-1801 Assumptions 2 Assumptions 3 Assumptions 4 Assumptions 5 Assumptions 6 Assumptions 7 n 257 N 257 n 257 N 257 Frequency range # of antenna in an antenna module/set (# of patches, # of dipoles, etc. ) # of antenna module/set in total Finite UV test points Beam phase shifter controller - - - - 1 1 2 2 2 3 3 Y/N Y Y Y Y degree 45 45 - - - - Antenna type (patch, dipole, or both) Antenna module/set location (front, back, top-side, left-side, right-side, bottom-side) Front cover (Plastic, Glass, Ceramic, Metal) Back cover (Plastic, Glass, Ceramic, Metal) Side cover / Frame (Plastic, Glass, Ceramic, Metal) Device size (Wx. Hx. D) Display panel – Full (Y) or Partial (N) Bezel Margin Top / Bottom Top & Bottom Notes Depends on the current implementation Finite test point shall be the baseline Finite beam shall be the baseline Depends on the current implementation Left & Right & combination of the lists are Bottom not precluded. Glass Glass Plastic Glass Plastic Metal Plastic cm 3 Y/N 66. 6 Y 66. 6 Y mm 1. 5 1. 5 This information is meaningful only if it’s the same with the material which covers antennas. This is for information Module can’t be placed outer edge of UE to secure mechanical reliability

Appendix: Work plan • Initiate offline and email discussion (after RAN 4#85) on the use cases and model assumptions for NW performance analysis • RAN 4 NR AH #4 (January ’ 18) • Initial discussion of simulation results (Both EIRP CDF and Network) based on the harmonized assumptions in this way forward. • Propose harmonized NW model assumptions and update based on preliminary analysis. • RAN 4 #86 (February ’ 18) • Deadline to submit the EIRP CDF simulation results based on the harmonized assumptions. Target preliminary EIRP CDF spherical requirement, based on the simulation outcomes. • Continue to improve the NW simulation accuracy reflecting initial EIRP CDF requirement (from AH #4) • Initial discussion of measurement results for prototype devices • RAN 4 #86 bis (April ’ 18) • Continue to improve the NW simulation accuracy reflecting preliminary EIRP CDF requirement (#86) • Continue to improve the prototype measurement effort and compare to preliminary EIRP CDF simulation • RAN 4 #87 (May ’ 18) • Finalize the spherical coverage requirement for handheld UEs based on the contributions