Coverage and Capacity Analysis of mmwave Cellular Systems

Size: px

Start display at page:

Download "Coverage and Capacity Analysis of mmwave Cellular Systems"

Melvyn Anderson
6 years ago
Views:

1 Coverage and Capacity Analysis of mmwave Cellular Systems Robert W. Heath Jr. The University of Texas at Austin Joint work with Tianyang Bai

competitive research ideas, vast expertise, first access to students Heavily funded center $5M About

2 Wireless is Big in Texas 20 Faculty 12 Industrial Affiliates Affiliates champion large federal proposals, provide technical input/feedback, unrestricted gift funds WNCG provides pre- prints, pre- competitive research ideas, vast expertise, first access to students Heavily funded center $5M About half of all students intern for an affiliate or work full- time 150 Grad Students Affiliates provide real world context 2

Wireless Systems Innovation Lab Sponsors champion large federal proposals, provide technical input/feedback, unrestricted gift

reviewed papers Several best paper awards Research dissemination through online videos Tight collaboration within WSIL Current

over wireless Cloud radio access networks Heterogeneous networks Massive MIMO Prototyping RF gesture recognition Applications of

competent interns Current and Past WSIL Sponsors Select Past Research Thrusts Interference alignment Multi-hop & relays

3 Wireless Systems Innovation Lab Sponsors champion large federal proposals, provide technical input/feedback, unrestricted gift funds 10 Full-time graduate students Wide collaboration inside and outside UT Comprehensive wireless expertise Hundreds of peer reviewed papers Several best paper awards Research dissemination through online videos Tight collaboration within WSIL Current Research Thrusts Millimeter wave cellular communication Millimeter wave for mobile ad hoc networks Perceptually-optimized video over wireless Cloud radio access networks Heterogeneous networks Massive MIMO Prototyping RF gesture recognition Applications of wireless in Oil and Gas WSIL provides pre-prints, precompetitive research ideas, vast expertise, first access to students, competent interns Current and Past WSIL Sponsors Select Past Research Thrusts Interference alignment Multi-hop & relays Viability of PHY research Limited feedback in MIMO systems Limited feedback in multiuser MIMO systems Limited feedback in multi cell MIMO systems 3

4 Wireless Communications Lab front.pdf 1 9/12/11 4:46 PM Undergrad/grad lab course QAM & OFDM experiments Complete lab manual & software Uses USRP equipment LabVIEW programming Complete lab manual available DIGITAL COMMUNICATIONS PHYSICAL LAYER EXPLORATION LAB USING THE NI USRP PLATFORM Dr. Robert W. Heath, University of Texas at Austin 4

5 Introduction

6 Why mmwave for Cellular? 1G-4G cellular 5G cellular Microwave m i l l i m e t e r w a v e 300 MHz 3 GHz 28 GHz GHz GHz 300 GHz Huge amount of spectrum available in mmwave bands* Cellular systems live with limited microwave spectrum ~ 600MHz 29GHz possibly available in 23GHz, LMDS, 38, 40, 46, 47, 49, and E-band Technology advances make mmwave possible Silicon-based technology enables low-cost highly-packed mmwave RFIC** Commercial products already available (or soon) for PAN and LAN Already deployed for backhaul in commercial products * Z. Pi,, and F. Khan. "An introduction to millimeter-wave mobile broadband systems." IEEE Communications Magazine, vol. 49, no. 6, pp , Jun ** T.S. Rappaport, J. N. Murdock, and F. Gutierrez. "State of the art in 60-GHz integrated circuits and systems for wireless communications." Proceedings of the IEEE, vol. 99, no. 8, pp: ,

The Need for Gain mmwave aperture mmwave noise bandwidth TX RX A eff = microwave aperture Smaller wavelength means smaller captured energy at antenna 3GHz->30GHz gives 20dB extra path loss due to

7 The Need for Gain mmwave aperture mmwave noise bandwidth TX RX A eff = microwave aperture Smaller wavelength means smaller captured energy at antenna 3GHz->30GHz gives 20dB extra path loss due to aperture microwave noise bandwidth P t P r = A eff 4 R = 2 4 R Larger bandwidth means higher noise power and lower SNR 50MHz -> 500MHz bandwidth gives 10dB extra noise power Solution: Exploit array gain from large antenna arrays 7

8 Antenna Arrays are Important highly directional MIMO transmission Baseband Processing Baseband Processing antennas are small (mm) ~100 antennas Narrow beams are a new feature of mmwave Reduces fading, multi-path, and interference Implemented in analog due to hardware constraints Arrays will change system design principles used at TX and RX 8

9 Traditional Beamforming Limitations Possible solution Precoding in the analog domain Bas eban RF RF Bas eban Power consumption limits the # of RF & ADC/DACs Analog beamforming has additional constraints Constant gains: Only phases is typically adjusted Quantized phases: Fixed set of steering directions are allowed Phase shifters Need beamforming strategies suitable for mmwave hardware 9

10 Hybrid Beamforming for mmwave Digital/ Baseband Analog/ RF Analog/ RF Digital/ Baseband RF Chain RF Chain N S Digital Precoder BS F BB N RF BS F RF N BS N MS BS F RF N RF Digital Precoder MS F BB N S RF Chain RF Chain Combine both digital and analog beamforming Small number of digital basebands (2 or 4) Allows more advanced MIMO strategies to be exploited Spatial multiplexing or multiuser MIMO Hybrid approach allows more advanced beam design O. El Ayach, S. Abu-Surra, S. Rajagopal, Z. Pi, and R. W. Heath, Jr., `` Spatially Sparse Precoding in Millimeter Wave MIMO Systems,'' submitted to IEEE Trans. on Wireless, May Available on ArXiv. 10

The mmwave Channel for Cellular LOS region d image of 38 and 60 GHz

mmwave cellular channel already measured in various environment Many

results validate the feasibility of mmwave cellular networks * Figures

Qiao, 38 GHz and 60 GHz angle dependent propagation for cellular &

11 The mmwave Channel for Cellular LOS region d image of 38 and 60 GHz peer-to-peer m mmwave cellular channel measurement at UT campus* mmwave cellular channel already measured in various environment Many characteristics of mmwave cellular channels are known Measurement results validate the feasibility of mmwave cellular networks * Figures from T. S. Rappaport, E. Ben-Dor, J. Murdock, Y. Qiao, 38 GHz and 60 GHz angle dependent propagation for cellular & peer-to-peer wireless communications, In Proc. of International Conference on Communications (ICC),

$exponent of LOS is 2 Diffraction is weaker in higher frequency, thus less loss when LOS NLOS (non-line-of-sight) is possible thanks to reflections Path loss$ exponent of NLOS (non-los) is about 4 Best reflected signal is still 20dB weaker than LOS * Figure from T. S. Rappaport, E. Ben-Dor, J. Murdock, Y.

exponent of NLOS (non-los) is about 4 Best reflected signal is still 20dB weaker than LOS * Figure from T. S. Rappaport, E. Ben-Dor, J. Murdock, Y.

12 Example Measurement Insights atter plot of the measured path loss values relative to 3 LOS (line-of-sight) signals propagate as in free space Path loss exponent of LOS is 2 Diffraction is weaker in higher frequency, thus less loss when LOS NLOS (non-line-of-sight) is possible thanks to reflections Path loss exponent of NLOS (non-los) is about 4 Best reflected signal is still 20dB weaker than LOS * Figure from T. S. Rappaport, E. Ben-Dor, J. Murdock, Y. Qiao, 38 GHz and 60 GHz angle dependent propagation for cellular & peer-to-peer wireless communications, In Proc. of International Conference on Communications (ICC),

13 Impact of Propagation Signal blocked by buildings Reflection Line-of-sight link Serving base station Blockages Blocked interfer Users may connect to a further unblocked base station Strong interferers may blocked Signal and interference may be either LOS or NLOS Need to include propagation models in the analysis T. Bai, V. Desai, and R. W. Heath Jr. Millimeter Wave Cellular Channel Models for System Evaluation, to appear in the Proc. of ICIN, Hawaii, Feb

mmwave Performance Analysis Directional Beamforming (BF) LOS & non-los

via main lobes to achieve array gain Steering directions at interfering

different characteristics in LOS & NLOS channels Better characterize

14 mmwave Performance Analysis Directional Beamforming (BF) LOS & non-los links Need to incorporate directional beamforming RX and TX communicate via main lobes to achieve array gain Steering directions at interfering BSs are random Need to distinguish LOS and NLOS paths Incorporate different characteristics in LOS & NLOS channels Better characterize blockages How to including beamforming + blockages in mmwave cellular analysis? 14

15 Stochastic Geometry for mmwave Cellular System Analysis

16 Stochastic Geometry for Cellular performance analyzed for a typical user base station locations distributed (usually) as a Poisson point process (PPP) Baccelli Stochastic geometry is a tool for analyzing microwave cellular Reasonable fit with real deployments Closed form solutions for coverage probability available Provides a system-wide performance characterization Need to incorporate LOS/non-LOS links and directional antennas J. G. Andrews, F. Baccelli, and R. K. Ganti, "A Tractable Approach to Coverage and Rate in Cellular Networks", IEEE Transactions on Communications, November T. X. Brown, "Cellular performance bounds via shotgun cellular systems," IEEE JSAC, vol.18, no.11, pp.2443,2455, Nov

17 Poisson Point Processes Antenna steering orientations as marks of the BS PPP Poisson point process (PPP): the simplest point process # of points is a Poisson variable with mean λs Given N points in certain area, locations independent Useful results like Campbell s Theorem & Displacement Theorem apply Assigning each point an i.i.d. random variable forms a marked PPP 17

18 Blockages in mmwave LOS: K=0 non-los K>0 Randomly located buildings Use random shape theory to model buildings Model random buildings as a rectangular Boolean scheme Buildings distributed as PPP with independent sizes & orientations Compute the LOS probability based on the building model # of blockages on a link is a Poisson random variable Boolean scheme of rectangles K: # of blockages on a link The LOS probability that no blockage on a link of length R is e R T. Bai and R. W. Heath, Jr., ``Using Random Shape Theory to Model Blockage in Random Cellular Networks,'' Proc. of the International Conf. on Signal Processing and Communications, Bangalore, India, July 22-25,

19 Directional Transmission at the BS 10 9 Exact antenna pattern "Sector" approximation Antenna Gain Half-Power BW Main lobe array gain M Main lobe beamwidth Angel in Rad Sector antenna approximation Back lobe gain m Each base station is marked with a directional antenna Antenna directions of interferers are uniformly distributed Use sector pattern in analysis for simplicity Antenna pattern fully characterized by, M and m 19

20 Proposed mmwave Model Buildings Serving BS Reflections Typical User PPP Interfering BSs Use stochastic geometry to model BSs as marked PPP Model the steering directions as independent marks of the BSs Use random shape theory to model buildings Model the building as rectangle Boolean schemes Different path loss exponents for LOS and non-los paths 20

21 System Parameters Different path loss model for LOS and nonlos links Line-of-sight with probability The fraction of land covered by buildings LOS path Loss in db: Non-LOS path loss in db: 28GHz system: let C=70 db, K=10 db General small scale fading : average LOS range is No fading case: small scaling fading is minor in mmwave [RapSun] Link budget Tx antenna input power: 30dBm Signal bandwidth: 500 MHz (Noise: -87 dbm) Noise figure: 5dB Z = Y e R 2 ([L]+[W ]) [W ][L] PL 1 = C + 20 log R(m) PL 2 = C + K + 40 log R(m) h 1/ Average building length and width 21

22 Results on Coverage

23 L i t < j2 t w. p. 2 1 M 0t Ak =,}, L = max {L t ) i : mt ZMw. M H `(r r t 0 0 p. 1 1 i 2 P sxmt sxmr sx sx P, SINR = L k um k u + (1 t + p (1 L = x ] = exp pt H pt )pr e pr )e umr N0 /P+t +Mr Mk>0 t pa r0e`(r t k B0k)H k `(rk ) r Mt, SINRM =8 xp Serving BS and User connect via main lobe i A Bk Hk `(rk ) 0 /Pt + sxmk>0 t mr r k <NM r Mt +(1 pt8 )(1 rpr )e w.ump. 2 1 (du), M M H `(r ) Bk = <, r t 0 0 x re : Cx, SINR =r 2 w. p. w. 8 m 1Pp.2 N /Pt + k>0 Ax k Bk H.k `(rk ) `(x) = < 20 i p. e CxKx 4L = w. x : C r, w. p. 1 e i i `(x) =. where General small-scale fading x : C Kx 4 w. p. 1 e 8 H0 `(r08 )< = min {H k `(rk )}, Connecting to the strongest signal before BF 2 x k>0 Cx w. p. e t < M w. p. 8 t `(x) =. Ak = <:L = max4 {L2, t}, x MC w. p. i w. p. 1 e : m t Kx t 2 w. i p. 1 t Ak = 2, Array gain of the TX antenna t : mt w. p t r < 8 M< w. p. M w. p. M M H `(r ) r tr t 2 P, Array SINR = Bk =A<, gain of the RX antenna =M/P, r kn w. p. Ak B krhk `(rk )t r t+ : m 0: k>0 2 w. p. 1 r mt w. p. Bk = 2 1, 2 r : mr w. p < Cx82 x w. p. e r < M {Hk `(r `(x)h=. Path Loss of LOS or non-los )}, w. p. 0 `(r0 ) = min k r 2 x := k>04 C Kx w. p. 1 e B, k H0 `(r0 ) = min {H `(r )}, k k r : SINR Expressions m k>0 r Use w. p stochastic<geometry to tcompute Mt w. p. 2 Ak =, : t SINR distribution 23

24 Z Coverage Probability of mmwave Theorem1 [mmwave Coverage probability] The coverage probability where P(SINR > T)= P[SINR >T] Z 1. 1 apple Z Z Z Z 1 f L (x) = Z f L (x) = Z apple Z d dx e (x) Z (x) =2 E h " Z ( xh K ) t 1 can be computed as U(x, t)f L (x) ej2 t/t 1 j2 t Z e t i dt + d, dxdt dx e (x) Z p xh T. Bai and R. W. Heath Jr., ``Coverage analysis of millimeter cellular networks with blockage effects (invited)", to appear in Proc. of IEEE GlobalSIP, Austin, TX, Dec te t dt #, 24

25 Coverage Gain from Large Arrays 1 Assume no RX beamforming Avg. cell redius: Rc=100 m Avg. LOS range: 1/ =141 m Microwave cell radius: 500 m Coverage Probability Gain from directional antenna array Gain from narrow beams Microwave: SU 4X4 MIMO 10 db gain, 30 beamwidth 20 db gain, 30 beamwidth 10 db gain, 60 beamwidth SINR threshold in db Large arrays provide better coverage probability Larger directivity gain provides better coverage Smaller beamwidth provides better coverage mmwave coverage probability comparable to microwave 25

26 Coverage Gain from Higher Density Tx directivity gain: 20 db Tx beamwidth: 30 degree Rx directivity gain: 10 db Rx beamwidth: 90 degree Avg. LOS range: 1/ =141 m Microwave cell radius: 500 m Coverage Probability Interference-limited region Noise-limited region due to high path loss Gain over microwave Microwave: SU 4X4 MIMO R c =50 m R c =100 m R c =200 m R c =300 m SINR threshold in db Higher density can also increase coverage probability Coverage probability no longer invariant with BS density Become interference-limited when coverage probability is good 26

27 LOS & non-los Path Loss 1 Tx directivity gain: 15 db Tx beamwidth: 30 degree Rx directivity gain: 10 db Rx beamwidth: 90 degree Rc=100 m 1/ =141 m Coverage Probability pure LOS (no buildings) Proposed model LOS path loss only proposed model Gain from blocking more interference 0.3 NLOS path loss only pure NLOS SINR threshold in db Coverage probability differs in LOS and non-los region Need to incorporate blockage model & differentiate LOS and nonlos Non-LOS coverage probability generally provides a lower bound Buildings may improve coverage by blocking more interference 27

28 Dense Network Analysis LOS Region of the typical user Equivalent LOS ball Typical User LOS BS Good coverage requires dense BS deployments LOS BSs exist with high probability in dense networks Noise and NLOS interference become much weaker than LOS interf. Theorem 1 sometimes inefficient to compute The underlying reason is that LOS region is very irregular Need to simplify expressions in Theorem 1 Approximate LOS region & neglect NLOS contributions 28

29 Insights for Dense Networks Theorem 2 [Coverage probability with dense BSs] P(SINR >T) e N X n=1 ( 1) n+1 Ǹ Z 1 0 4Y k=1 e a k(e nb k t te nb k ) 1 e n b k t 1 e n b k n ak b k t dt Given antenna patterns, SINR only depends on The larger, the denser the BSs (and closer) Increasing BS density need not improve SINR = Optimal BS density is finite Average Z Ysize of LOS region Average cell size 1 SINR goes to zero in a infinitely dense network p. SINR =0 lim 1 Y 2 2 What is the optimal BS density in dense networks? 29

30 Finding Optimal BS Density Tx directivity gain: 20 db Tx beamwidth: 30 degree Rx directivity gain: 10 db Rx beamwidth: 10 db Avg. LOS range: 1/ =141 m Target SINR: T=10 db Coverage Probability with T=10 db Increasing BS density need not improve SINR Exhuastive search using Simulations Exhaustive search using Theorem 2 Optimal BS density is finite Average cell radius Exhaustive search optimal BS density using Theorem 2 Maximize the coverage probability given a target SINR Much efficient than simulations with minor errors Optimal cell radius is approximately 2/3 of the avg. LOS range. 30

31 Finding Optimal BS Density Tx directivity gain: 20 db Tx beamwidth: 30 degree Rx directivity gain: 10 db Rx beamwidth: 10 db Avg. LOS range: 1/ =141 m Target SINR: T=20 db Coverage Probability Simulations Analytical Results Optimal BS density is finite Gap by ignoring NLOS and noise Increasing BS density need 0.2 not improve SINR Avg. Cell Radius R c Exhaustive search optimal BS density using Theorem 2 Maximize the coverage probability given a target SINR Much efficient than simulations with minor errors Optimal cell radius is approximately 2/3 of the avg. LOS range. 31

32 Results on Rate

33 Data Rate Comparison Given coverage probability, the achievable rate is Microwave network 4X4 SU MIMO with bandwidth 50MHz: Spectrum efficiency is 4.56 bps/ Hz Data rate is 226 Mbps (Rc=500 m) mmwave network with bandwidth 500MHz: Tx beamwidth: 30 degree Rx directivity gain: 10 db Rx beamwidth: 90 degree Avg. LOS range: 1/ =141 m R = 1 ln(2) R c M 100m 200m 10 db 2.76 Gbps 1.61 Gbps 20 db 2.91Gbps 1.88 Gbps Tx beamforming Gain Z C mmwave achieves high gain in average rate 0 P c (T ) 1+T dt clipped by 6 bps/hz (64QAM) Average cell radius 33

34 SINR Coverage Comparison 1 SINR Coverage Probability mmwave: 30 db BF gain, R c =100 m Massive MIMO: N =, 8 Users/ BS t CoMP: N =4, 2 Users/ BS, 3 BSs/ Cluster t SU MIMO: 4X4 Gain from directional antennas and blockages in mmwave Gain from larger number of antennas SINR Threhold in db SINR CCDF 34

35 Spectrum Efficiency Comparison Spectrum efficiency for single user Gain from directional antennas and blockages in mmwave Rate Coverage Probability mmwave: 30 db BF gain, R c =100 m Massive MIMO: N t =, 8 Users/ BS CoMP: N t =4, 2 Users/ BS, 3 BSs/ Cluster SU MIMO: 4X4 Gain from massive number of antennas Spectrum Efficiency (bps/hz) mmwave can support much higher spectral efficiency * for more information on this setup refer to: Robert W. Heath Jr., Role of MIMO Beyond LTE: Massive? Coordinated? mmwave?, Workshop on Beyond 3GPP LTE-A ICC

36 Cell Throughput Comparison Gain from larger bandwidth Rate Coverage Probability mmwave: 30 db BF gain, R c =100 m Massive MIMO: N t =, 8 Users/ BS CoMP: N t =4, 2 Users/ BS, 3 BSs/ Cluster SU MIMO: 4X4 ZF Gain from serving multiple users Cell Throughput in Mbps mmwave can support much higher data rate * for more information on this setup refer to: Robert W. Heath Jr., Role of MIMO Beyond LTE: Massive? Coordinated? mmwave?, Workshop on Beyond 3GPP LTE-A ICC

37 Conclusions

38 Going Forward with mmwave mmwave coverage probability and rate Need to include both LOS and Non-LOS conditions Interference is reduced by directional antennas and blockages Good rates and coverage can be achieved Theoretical challenges abound Analog beamforming algorithms & hybrid beamforming Channel estimation, exploiting sparsity, incorporating robustness Multi-user beamforming algorithms and analysis Microwave-overlaid mmwave system a.k.a. phantom cells Going away from cells to a more ad hoc configuration 38

39 Questions? T. Rappaport, R. W. Heath Jr., R. C. Daniels, and J. Murdock, Millimeter Wave Wireless Communications, Prentice Hall, (expected) O. El Ayach, S. Abu-Surra, S. Rajagopal, Z. Pi, and R. W. Heath, Jr., `` Spatially Sparse Precoding in Millimeter Wave MIMO Systems,'' submitted to IEEE Trans. on Wireless, May Available on ArXiv. N. Valliappan, A. Lozano, and R. W. Heath, Jr.``Antenna Subset Modulation for Secure Millimeter-Wave Wireless Communication,'' to appear in IEEE Trans. on Communications. R. Daniels, J. N. Murdock, T. Rappapport, and R. W. Heath, Jr., ``State-of-the-Art in 60 GHz,'' IEEE Microwave Magazine, vol. 11, no. 7, pp , Dec Cheol Hee Park, R. W. Heath, Jr., and T. Rappapport, ``Frequency-Domain Channel Estimation and Equalization for Continuous-Phase Modulations with Superimposed Pilot Sequences,'' IEEE Trans. on Veh. Tech., vol. 58, no. 9, Nov R. Daniels and R. W. Heath, Jr., ``60 GHz Wireless Communications: Emerging Requirements and Design Recommendations,'' IEEE Vehicular Technology Magazine, vol. 2, no. 3, pp , Sept T. Bai, V. Desai, and R. W. Heath Jr. Millimeter Wave Cellular Channel Models for System Evaluation, to appear in the Proc. of ICIN, Hawaii, Feb T. Bai and R. W. Heath, Jr., ``Using Random Shape Theory to Model Blockage in Random Cellular Networks,'' Proc. of the International Conf. on Signal Processing and Communications, Bangalore, India, July 22-25, S. Akoum, O. El Ayach and R. W. Heath, Jr., `` Coverage and Capacity in mmwave MIMO Systems,'' (invited) in Proc. of the IEEE Asilomar Conf. on Signals, Systems, and Computers, Pacific Grove, CA, Nov. 4-7, T. Bai and R. W. Heath Jr., ``Coverage analysis of millimeter cellular networks with blockage effects (invited)", to appear in Proc. of IEEE GlobalSIP, Austin, TX, Dec

Millimeter Wave Cellular Channel Models for System Evaluation

Millimeter Wave Cellular Channel Models for System Evaluation Tianyang Bai 1, Vipul Desai 2, and Robert W. Heath, Jr. 1 1 ECE Department, The University of Texas at Austin, Austin, TX 2 Huawei Technologies,