Tsoulos George (ред.) MIMO System Technology for Wireless Communications
Подождите немного. Документ загружается.
364 MIMO System Technology for Wireless Communications
12.2 Deployment
MIMO capacity can show large local variations, depending on the position
of the antenna arrays and on the environment. How does one deploy MIMO
systems to maximize capacity practically? It is concluded from double-
directional MIMO channel measurements [19] that directions of arrival and
directions of departure may be discrete. Such environments lend themselves
for transmission into limited angles. Even if the directions of departure and
arrival are not discrete, it is likely that there are specific sectors from which
much power can be received. This favors systems with directional transmis-
sion. The measurements in office environments [7,19] clearly demonstrate
that there exist at least large angular sectors, either from which no significant
power was received or into which no power is transmitted, that would
eventually reach an intended receive antenna. How does the terminal or the
access point know which are the favored directions? The deployment pro-
cedure proposed in [20] relies on measurements in the actual propagation
environment.
Transmit diversity schemes are particularly appropriate for microcell sce-
narios [4] where capacity gains are large and beamforming enhancements are
less appropriate due to the large angular spread [21]. Transmit diversity
provides a relatively simple capacity upgrade solution in terms of configur-
ing additional hardware. For the Alamouti scheme, two transmit antennas are
required on the downlink — either a single cross-planar antenna or a pair
of vertically polarized antennas. In terms of power amplifier modules, the
operator may be able to share existing power. This is possible when the site
is configured with multiple carriers and multi-carrier power amplifiers. This
is not possible for single-carrier scenarios, and supplementary power ampli-
fiers must be included.
12.3 Smart Antenna Planning Example
In this section, we discuss a specific planning problem through a simulation
study. We discuss the connectivity of users in a 3G cellular network with
conventional sectorization and compare it with the connectivity when the
network deploys 4-element uniform linear array antennas.
Specifically, we assume a UMTS radio access network configuration
including approximately 100 sectors. In order to investigate the system per-
formance, system-level simulations were carried out using a state-of-the-art
Monte Carlo simulator. The system-level simulator models the network sim-
ilarly to most of the commercially available radio network planning tools.
Additionally, we included the modeling of smart antenna processing. In this
4190_book.fm Page 364 Tuesday, February 21, 2006 9:14 AM
Network Planning and Deployment Issues for MIMO Systems 365
study, we consider two deployment scenarios: In the first scenario, we con-
sider all base stations to be equipped with conventional sector antennas (three
sectors per site). In the second scenario, the base stations are equipped with
4-element uniform linear array antennas. In the second scenario, we assume
that the baseband signal processing at the base station implements optimum
combining.
The typical questions to be answered by the operator are:
• How do smart antennas influence the system performance?
• In which regions are (dis-)connected users concentrated?
• Does the distribution of (dis-)connected users change in the case of
smart antenna deployment?
• Does it make sense to deploy smart antennas throughout the entire
network or only in certain areas?
For simplicity we assume an idealized service mix with just two different
service types, i.e., we assume 40% of the users to be served at a data rate of
12.2 kbit/s and 60% of the users at 64 kbit/s.
The results are shown in Figure 12.6 and Figure 12.7. Figure 12.6 shows
the distribution of the connected users for conventional sector antennas (a)
and 4-element-ULAs (b). Conversely, Figure 12.7 shows the distribution of
disconnected users in case of conventional sector antennas (a) and 4-element-
ULAs (b).
This comparison shows that smart antennas are capable of increasing the
number of served users. Hence, the network’s aggregated throughput can
be increased substantially by the use of smart antennas. In the studied
scenario, the deployment of 4-element-ULAs is capable of doubling the
network throughput. Only a very small number of users remain discon-
nected in areas of poor coverage.
We note that smart antennas provide a “cluster gain,” rather than just a
performance gain limited to the smart antenna serving cell. This means that
a single cell equipped with a smart antenna will not just improve the cov-
erage and capacity of the serving sector, but will also improve the perfor-
mance of the neighbouring cells, due to the highly reduced intercell
interference power.
Recently, a statistical analysis in live CDMA networks was carried out in
North America [22]. The analysis has shown that among all sites in which
the load exceeds the capacity threshold, around 51% have a single over-
loaded sector, while the other sectors operate below the capacity threshold.
Another 38% of the sites have two sectors exceeding the capacity threshold,
and only 11% of the sites have all three sectors exceeding the capacity
threshold. This means that, when deployed per sector, smart antennas allow
the operator to address capacity where additional resources are needed,
increasing the utilization of every deployed carrier.
4190_book.fm Page 365 Tuesday, February 21, 2006 9:14 AM
366 MIMO System Technology for Wireless Communications
(a)
(b)
FIGURE 12.6
Distribution of connected users in the system area. Each base station is either equipped with a
single element sector antenna (a) or a 4-element uniform linear array (ULA, performing opti-
mum combining) (b), respectively. The “.” indicates users with a 12.2 kbit/s service, and the
“x” shows users with 64 kbit/s service.
Distance(km)
Distance(km)
1
2
3
4
5
6
7
8
9
10
11
1 2 3 4 5 6 7 8 9 10 11
Distance(km)
1
2
3
4
5
6
7
8
9
10
11
Distance(km)
1 2 3 4 5 6 7 8 9 10 11
4190_book.fm Page 366 Tuesday, February 21, 2006 9:14 AM
Network Planning and Deployment Issues for MIMO Systems 367
(a)
(b)
FIGURE 12.7
Distribution of disconnected users in the system area. Each base station is either equipped with
a single element sector antenna (a), or a 4-element uniform linear array (ULA, performing
optimum combining) (b), respectively. The “.” indicates users with a 12.2 kbit/s service, and
the “x” shows users with 64 kbit/s service.
Distance(km)
1
2
3
4
5
6
7
8
9
10
11
Distance(km)
1 2 3 4 5 6 7 8 9 10 11
Distance(km)
1
2
3
4
5
6
7
8
9
10
11
Distance(km)
1
2
3
4
5 6
7
8
9
10
11
4190_book.fm Page 367 Tuesday, February 21, 2006 9:14 AM
368 MIMO System Technology for Wireless Communications
Considering both the sector-based capacity limitations and the cluster gain
of smart antennas, future challenges for smart antenna and MIMO radio
network deployment will be to find answers to the following questions:
• How many smart antennas/MIMO systems should be deployed?
• Where should they be deployed?
• How should the mixed deployment of conventional sector antennas,
smart antennas, and MIMO systems be configured in order to pro-
vide the best system performance?
References
1. W. Weichselberger. 2003. “Spatial Structure of Multiple Antenna Radio Chan-
nels — A Signal Processor Viewpoint,” Doctoral Dissertation, Vienna Univer-
sity of Technology, Vienna, Austria. Available at: http://www.nt.tuwien.ac.at/
mobile/theses_finished.
2. S.N. Diggavi, N. Al-Dhahir, A. Stamoulis, and A.R. Calderbank. 2004. “Great
expectations: the value of spatial diversity in wireless networks,” Proceedings
of the IEEE, Vol. 92, No. 2, Feb. 2004, pp. 219–270.
3. A. Paulraj, R. Nabar, and D. Gore. Introduction to Space-Time Wireless Commu-
nications, Cambridge, U.K., Cambridge University Press.
4. J. Laiho, A. Wacker, and T. Novosad, eds. 2002. Radio Network Planning and
Optimization for UMTS, New York, John Wiley & Sons.
5. A. Kuchar. 1999. “Real Time Smart Antenna Processing for GSM1800”, Doctoral
Dissertation, Vienna University of Technology, Vienna, Austria, 1999. Available
at: http://www.nt.tuwien.ac.at/mobile/theses_finished.
6. Towards Mobile Broadband Multimedia Networks, Cooperation euroéenne
dans le domaine de la recherché Scientific et Technique Action 273. Available
at: http://www.lx.it.pt/cost273/.
7. H. Özcelik, M. Herdin, W. Weichselberger, J. Wallace, and E. Bonek. 2003.
“Deficiencies of ‘Kronecker’ MIMO radio channel model,” Electronics Letters,
Vol. 39, No. 16, pp. 1209–1210.
8. Durlacher Research Ltd. 2001. “UMTS-Report — An Investment Perspective,”
Press Release. Available at: http://www.dad.be/library/pdf/durlacher3.pdf.
9. 3GPP TS 32.642. 2005. Telecommunication management; Configuration Man-
agement (CM); UTRAN network resources Integration Reference Point (IRP):
Network Resource Model (NRM), Release 6, Version 6.4.0, April 2005.
10. T. Giles, J. Markendahl, J. Zander, P. Zetterberg, P. Karlsson, J. Lind, and
G. Malmgren. 2004. “Cost drivers and deployment scenarios for future broad-
band wireless networks — key research problems and directions for research,”
Proc. IEEE Vehicular Technology Conference, Spring, May 2004.
11. M. Haardt, C.F. Mecklenbräuker, M. Vollmer, and P. Slanina. 2001. Smart An-
tennas for UTRA TDD, European Transactions on Telecommunications (ETT),
Special Issue on Smart Antennas, Vol. 12, Issue 5, pp. 393–406, Sept.–Oct. 2001.
12. ETSI TR 101 112. 1998. “Selection procedures for the choice of radio transmis-
sion technologies of the universal mobile telecommunications system UMTS
(UMTS 30.03), Version 3.2.0,” April 1998.
4190_book.fm Page 368 Tuesday, February 21, 2006 9:14 AM
Network Planning and Deployment Issues for MIMO Systems 369
13. 3GPP TS 25.211. 2005. Technical Specification Group Radio Access Network;
Physical channels and mapping of transport channels onto physical channels
(FDD) Release 6, Version 6.4.0, March 2005.
14. 3GPP TS 25.214. 2005. Technical Specification Group Radio Access Network;
Physical layer procedures (FDD), Release 6, Version 6.5.0, March 2005.
15. 3GPP TR 25.876. 2005. Technical Specification Group Radio Access Network;
Multiple-Input Multiple Output in UTRA, Release 7, Version 1.8.0, October
2005.
16. S. Catreux, P.F. Driessen, and L.J. Greenstein. 2000. “Simulation results for an
interference-limited multiple-input multiple-output cellular system,” IEEE
Comm. Lett., Vol. 4, No. 11, Nov. 2000, pp. 334–336.
17. S. Catreux, P.F. Driessen, and L. Greenstein. 2001. “Link-optimal space-time
processing with multiple transmit and receive antennas,” IEEE Comm. Lett.,
Vol. 5, pp. 334–336.
18. H. Dai, A.F. Molisch, and H.V. Poor. 2004. “Downlink capacity of interference-
limited MIMO systems with joint detection,“ IEEE Trans. Wireless Comm., Vol. 3,
pp. 442–453.
19. M. Steinbauer et al. 2001. “The double-directional radio propagation channel,”
IEEE Antennas & Propagation Magazine, Vol. 43, No. 4, August 2001, pp. 51–63.
20. E. Bonek. 2002. “Link-specific MIMO system deployment for ad-hoc networks,”
Proc. WWRF Meeting #7, Working Group 4, Eindhoven, the Netherlands, Decem-
ber 3–4, 2002.
21. S. Andersson et al. 1999. “Adaptive antennas for GSM and TDMA systems.”
IEEE Personal Communications, June 1999, pp. 74–86.
22. T. Crook. 2002. “Why smart antennas & subscriber device based enhancements
are important to wireless service providers,” Sprint. Available at: http://www.
cdg.org/news/events/CDMASeminar/cdg_tech_forum_02/1_sprint_cdg_
smart_antenna_rev.pdf.
4190_book.fm Page 369 Tuesday, February 21, 2006 9:14 AM
371
Index
A
A/D converter, 280–281
Adaptive modulation and coding, 213
Additive white Gaussian noise, 61, 91, 96,
116, 182, 230
Alamouti's transmit technique, 94–96
Algorithms
coordinated zero-forcing, 189–190
quantized signal adaptation, 121–133
V-BLAST, 90, 105–107, 230–231
Viterbi, 100–102
Analog beamforming, 168
Angle-of-arrival
description of, 3, 5
Gaussian, 11
Angular shift, 53–54
Angular spread, 11–12
Antenna selection
capacity-achieving spatial multiplexing,
162–165
description of, 148–149
diversity, 149
implementing of, 161–166
interference and, 154
multifunctional reconfigurable antenna,
See Multifunctional reconfigurable
antenna
multifunctional reconfigurable spiral
antenna, See Multifunctional
reconfigurable spiral antenna
non-idealities, 166–167
with orthogonal space–time block codes,
157–159
performance of, 152–154
receive, 148
smart antenna, 364–368
space–time codes, 165–166
with space–time trellis codes, 155–157
with spatial correlation, 167–170
summary of, 170–171
transmit
description of, 148
optimum signaling with, 151–152
transmit and receive, 148
Array gain, 149, 354
Automatic repeat request, 355
B
Base stations, 360, 362–363
Beamforming
base stations with, 362–363
description of, 121–122
fixed, 215
general coordinated, 190
interpolated, in MIMO-OFDM systems,
136–137
multiuser, 239–244
space–time coding and, 256
specific, 215
universal terrestrial radio access
frequency division duplex, 215–216
Bit error rates, 344
Block-fading channel assumption, 166
Broadband signaling, 113–114
Butler matrix, 170
C
Capacity
MIMO, See MIMO capacity
of multiple-user MIMO channels, 181–183
Capacity-achieving spatial multiplexing,
162–165
CapEx, 360
Channel estimation, for MIMO-OFDM,
324–330, 348–349
Channel fading, 92
Channel inversion
description of, 183–184
regularized, 184, 186
Channel memory, 115–116
Channel quality indicator, 213
Channel quantization, 119–120
4190_Idx.fm Page 371 Monday, February 20, 2006 10:31 AM
372 MIMO System Technology for Wireless Communications
Channel state information
description of, 92
space–time coding with, 107–109
at transmitter, 34–35, 180
Closed-loop MIMO systems, 116
Codebook
description of, 119
for spatial multiplexing, 131–133
Common pilot channels, 212
Coordinated zero-forcing, 189–190
Cumulative distribution function, 47, 153,
242–243
D
Dedicated physical channel, 211
Dedicated physical control channel, 211, 219,
363
Dedicated physical data channel, 211, 219
Degrees of freedom
description of, 38–39
effective, 306
Delay spread, 11–12
Digital signal processors, 277, 317, 330
Digital-intermediate frequency, 280
Direct channel quantization, 117
Directional channel modeling
angular spread, 11–12
delay spread, 11–12
extended Saleh-Valenzuela model, 10
Gaussian scatter density model, 10–11
Gaussian wide sense stationary
uncorrelated scattering, 7–8
geometrically based single-bounce
statistical channel models
description of, 5
elliptical model with dense discrete
scatterers, 7
elliptical subregions model, 7
geometrically based circular model, 5
geometrically based elliptical model, 6
overview of, 2–3
scatterers, ring of, 4
semi-elliptical geometrical model, 11
shadow fading, 11–12
stochastic spatio-temporal propagation
model, 8–10
Dirty-paper coding, 182, 191, 233
Discrete uniform distribution model, 4–5
Discrete-time simulation models, for phase
noise, 289–290
Diversity gains, 354–355
Diversity MIMO, 218–219
Diversity order, 149
Doppler shift, 3
Double Poisson time-of-arrival model, 10
Double transmit antenna array, 237
Downlink channel, 177
DSTTD-SGRC, 235–236
E
Effective degrees of freedom, 306
Electromagnetism, 59–60
Elliptical subregions model, 7
Ergodic capacity, 40
Euclidean distance, 155
Extended Saleh-Valenzuela model, 10
F
Feedback
in broadband channels with
MIMO-OFDM
description of, 135
limited feedback, 135–136
continuous quality improvement, 231
limited
description of, 119–121
MIMO-OFDM system and, 135–136
overview of, 113–114
quantized signal adaptation
algorithms, 121–133
description of, 120–121
spatial correlation effect, 133–135
Fermat principle, 13
5.25 GHz 2
× 2 MIMO testbed, 290–292
Fixed beamforming, 215
Flat-fading assumption, 166
Foschini-Telatar formula, 58–59
Frequency offset calibration, 295–297
Functional mapping, 321
G
Gaussian scatter density model, 10–11
Gaussian wide sense stationary uncorrelated
scattering, 7–8
General capacity formula, 31–32
4190_Idx.fm Page 372 Monday, February 20, 2006 10:31 AM
Index 373
General coordinated beamforming, 190
Geometrical theory of diffraction, 13
Geometrically based circular model, 5
Geometrically based single-bounce statistical
channel models
description of, 5
elliptical model with dense discrete
scatterers, 7
elliptical subregions model, 7
geometrically based circular model, 5
geometrically based elliptical model, 6
Grassmann manifold, 127
H
High speed downlink packet access, 212–214
Hybrid-selection maximum ratio combining,
159
I
Ideal scattering, 58
Imaging technique, 14
Implementation loss, 280
Information MIMO, 209, 219–220
Intelligence, 252
Interference
antenna selection and, 154
inter-symbol, 293
multiple access, 220–221
as noise, 182–183
Interference cancellation, 220
Interpolated beamforming, in MIMO-OFDM
systems, 136–137
Interpolated precoding, in MIMO-OFDM
systems, 138
Inter-symbol interference, 293
I/Q mismatch
description of, 282–284
simulation model for, 284
at transmitter, 283
K
Keyhole channel effect, 255
K-factor, 256
Kronecker model, 167
Kronecker symbol, 327
L
Laplacian power azimuth spectrum, 11
Layer space–time, 93, 104
Limited feedback
description of, 119–121
MIMO-OFDM system and, 135–136
Linear precoders, 184–187
Linear processing, multi-antenna receivers,
187–191
Linear processing, single-antenna receivers,
183–187
Linear receiver inspired selection, of
precoder, 129
Linear receivers, 164–165
Local field principle, 13
Local oscillators
description of, 280
low phase noise, 286–288
Low phase noise local oscillators, 286–288
M
Macrocell model, 5
Maximal ratio combining, 216
Maximum likelihood detector, 155
Maximum likelihood receiver inspired
selection, of precoder, 129
Maximum ratio combining, 159
Metal-insulator-metal capacitors, 261
Metropolitan networks, 355
Microcell model, 6
Microelectromechanical systems
description of, 250
radiofrequency
description of, 250
integrated antennas, 258–268
microwave laminates for MRA
fabrication, 257–258
three-element selection diversity
antenna, 258–263
MIMO capacity
definition of, 61
description of, 36–37, 58–59, 90–92
diversity, 92–93
extended capacity formula, 31–35
Foschini-Telatar formula for, 58–59
for Rayleigh channels
description of, 42–43
with spatial fading correlation, 49–50
without spatial fading correlation,
48–49
4190_Idx.fm Page 373 Monday, February 20, 2006 10:31 AM