An Overview of MIMO Wireless Systems

CHAPTER 2

An Overview of MIMO Wireless Systems

2.1 Introduction
In recent years, wireless communication devices have become more and more
prevalent since they allow users to communicate with each other independently from their
locations. At the same time, the demand for high data rate and reliable radio systems has
grown. Unfortunately, the design of systems satisfying these requirements is difficult in
traditional wireless channel using one antenna at the transmitter and at the receiver. Indeed,
due to the fading caused by multiple reflections of the transmitted signal, the reliability of
these systems is not guaranteed during the communication. In addition, the capacity of such
systems is limited. To handle with these limitations, the use of multiple antennas at both sides
is a promising solution. In fact, increasing the number of antennas leads to important benefits
regarding the reliability and the transmission throughput.
This chapter gives an overview wireless channel as well as the multiple antenna systems. It
also introduces their important parameters which are necessary for the next chapters.

2.2 The Wireless Channel

The wireless channel is the channel that uses radio waves to transmit data between
two devices, namely the transmitter and the receiver. It is characterized by its random time-
variation variable, called fading. At the receiver, the channel fading is seen as a multiplicative
process that affects the transmitted signal. In practice, the radio signal reaches the receiver via
different paths, A: Free space, B: Reflection, C: Diffraction, D: Scattering, as shown in
Figure 2-1. These different paths are said to be multi-path that experience differences in
attenuation, delay and phase shift. The superposition of these paths results in fading. Due to
6
 An Overview of MIMO Wireless Systems

the propagation delay, the fading is presented as a complex number with a phase and a
magnitude that represents the channel distortion. Different statistical models are associated to
the fading, depending on the propagation environment [26, 27].

Figure 2-1 Multipath Fading Channel.

Very often, especially in mobile communications, not only do multiple propagation paths
exist, but they are also time-varying. The result is a time-varying fading channel.
Communication through these channels can be difficult. Special techniques may be required
to achieve satisfactory performance.
The general time varying fading channel model is too complex for understanding and
performance analysis for wireless channels. One approximate channel model is the wide-
sense stationary uncorrelated scattering (WSSUS). In WSSUS model, the time-varying fading
process is assumed to be wide-sense stationary random process and the signal copies from the
scatterings by different objects are assumed to be independent.
The following parameters are often used to characterize a WSSUS channel:
1) Multipath Spread Tm
It tells us the maximum delay between paths of significant power in the channel.
2) Coherence Bandwidth (f)c
It gives an idea of how far apart in frequency for signals to undergo different degrees of
fading.
3) Coherence Time (t)c

7
 An Overview of MIMO Wireless Systems

It gives a measure of the time duration over which the channel impulse response is
essentially invariant (highly correlated)
4) Doppler Spread Bd
It gives the maximum range of Doppler shifts

2. 2.1Time Selective Fading, Frequency Selective Fading

Fading is a time varying effect. This variation is characterized by the channel
coherence time (t)c that serves as a measure of how fast the channel changes in time. In fact,
the coherence time corresponds to the longest interval during which the channel is assumed to
be constant. Let’s denote by Ts the time needed to transmit a symbol.

Using these parameters, different channel models could be distinguished:

 Ergodic: When (t)c = Ts, the channel is said to be Ergodic. In this case, each
transmitted symbol is associated with a new realization of the channel.
 Quasi-static: If the channel remains constant during one frame, i.e. (t)c = Tf , the
channel is called quasi-static.
 Block fading: When fading does not change during N frame transmissions, the
channel is said to be block fading. In this case, (t)c = q*Tf , Tf is the frame time. It is
noted that a quasi static channel is also a block fading channel with q = 1.

In the wireless channel, the multi-path propagation leads to a time delay spread to receive the
different paths. This spread is characterized in the frequency domain by the coherence
bandwidth, (f)c. Indeed, (f)c is a measure of the range of frequencies over which the
channel has approximately equal gain.
 When the signal bandwidth is comparable or less than (f)c, all the frequency
components of the transmitted signal undergo the same attenuation, the channel is
said to be flat or non-frequency selective fading.
 When the radio channel has different gains within the signal bandwidth, a frequency
selective fading is experienced.

8
 An Overview of MIMO Wireless Systems

2.2.2 Statistical Model for Fading Channels

There are several distributions that can be considered in attempting to model the
statistical characteristics of the fading channels. Some are:

 Rayleigh fading channels

 Rician fading channels
 Nakagami fading channels
 Lognormal fading channels

The Rayleigh distribution is frequently used to model the statistics of signals transmitted
through radio channels such as mobile channels. Assume that there are many objects in the
environment that scatter the radio signal before it arrives at the receiver. When there is large-
number of paths, applying Central Limit Theorem, each path can be modeled as a circularly
symmetric complex Gaussian random variable with time as the variable. This model is
called Rayleigh fading channel model. For more information, [28] presents and analyzes
Rayleigh fading model.

A circularly symmetric complex Gaussian random variable is of the form,

r=x+jy (2.1)

where the real and the imaginary parts of r are zero mean independent and identically
distributed (i.i.d.) Gaussian random variables.

The magnitude |r| which has a probability density,

𝑟 −𝑟 2
𝑝 𝑟 = 𝜎 2 exp , 𝑟≥0 (2.2)
𝑟 2𝜎𝑟2

Where,

r is called a Rayleigh random variable.

𝜎𝑟2 = E(r2) is the mean received power.

This model, called Rayleigh fading channel model, is reasonable for an environment where
there is large number of reflectors.

9
 An Overview of MIMO Wireless Systems

2.3 Multi-antenna System

Different antenna configurations used in defining space-time systems are illustrated in
Figure 2-2. Single-input single-output (SISO) is the well-known wireless configurations,
single-input multiple-output (SIMO) uses a single transmitting antenna and multiple (Nr)
receive antennas, multiple-input single-output (MISO) has multiple (Nt) transmit antennas
and one receive antenna and, finally MIMO has multiple (Nt) transmit antennas and multiple
(Nr) receive antennas [29].

Tx Rx Rx
Tx

SISO MISO

Tx
Rx Tx Rx

SIMO MIMO

Figure 2-2 Different antenna configurations in space-time systems.

2.4 MIMO Performance Gains

Wireless communications systems are increasingly under pressure to provide higher
data rates at superior levels of quality of service. Wireless communication systems employing
multi-antenna configurations have the ability to provide additional performance gains, not
available to traditional single antenna transceiver configurations. These additional gains may
be classified as either (i) array gain, (ii) diversity gain, or (iii) spatial multiplexing gain and
each may be used to improve link performance [9,29].

2.4.1 Array Gain

Due to the use of multiple antennas, the antenna gain is increased and this leads to an
increased range and coverage. This is useful in remote areas with low population. A large
area can thus be served with less base-stations. Alternatively, the transmit power of the
mobile units can be reduced due to the increased gain, or sensitivity, of the receiving base-
station antenna array.
10
 An Overview of MIMO Wireless Systems

2.4.2 Diversity Gain

Multipath fading is a significant problem in communications. In a fading channel,
signals experience fades (i.e., they fluctuate in their strength). When the signal power drops
significantly, the channel is said to be in a fade. This gives rise to high bit error rates (BER).
In this case, diversity is used to combat fading [27]. This involves providing replicas of the
transmitted signal over time, frequency, or space. There are three types of diversity schemes
in wireless communications.

 Temporal diversity: In this case replicas of the transmitted signal are provided across
time by a combination of channel coding and time interleaving strategies. The key
requirement here for this form of diversity to be effective is that the channel must
provide sufficient variations in time. It is applicable in cases where the coherence time
of the channel is small compared with the desired interleaving symbol duration. In
such an event, the interleaved symbol is independent of the previous symbol. This
makes it a completely new replica of the original symbol.
 Frequency diversity: This type of diversity provides replicas of the original signal in
the frequency domain. This is applicable in cases where the coherence bandwidth of
the channel is small compared with the bandwidth of the signal. This assures that
different parts of the relevant spectrum will suffer independent fades.
 Spatial diversity: This is also called antenna diversity and is an effective method for
combating multipath fading. In this case, replicas of the same transmitted signal are
provided across different antennas of the receiver. This is applicable in cases where
the antenna spacing is larger than the coherence distance to ensure independent fades
across different antennas. The traditional types of diversity schemes are [22] selection
diversity, maximal ratio diversity, and equal gain diversity.

Basically the effectiveness of any diversity scheme lies in the fact that the receiver
must provide independent samples of the basic signal that was transmitted. In such an event,
the probability of two or more relevant parts of the signal undergoing deep fades will be very
small. The constraints on coherence time, coherence bandwidth, and coherence distance
ensure this. The diversity scheme must then optimally combine the received diversified
waveforms so as to maximize the resulting signal quality. Diversity can also be categorized

11
 An Overview of MIMO Wireless Systems

under the subheading of spatial diversity, based on whether diversity is applied to the
transmitter or to the receiver.
 Receive diversity: Maximum ratio combining is a frequently applied diversity scheme
in receivers to improve signal quality. In cell phones it becomes costly and
cumbersome to deploy. This is one of the main reasons transmit diversity became
popular, since transmit diversity is easier to implement at the base station.
 Transmit diversity: In this case, controlled redundancies are introduced at the
transmitter, which can be then exploited by appropriate signal processing techniques
at the receiver. Generally this technique requires complete channel information at the
transmitter to make this possible. But with the advent of space-time coding schemes
like Alamouti’s scheme [3], it became possible to implement transmit diversity
without knowledge of the channel. This was one of the fundamental reasons why the
MIMO industry began to rise. Space-time codes for MIMO exploit both transmit as
well as receive diversity schemes, yielding a high quality of reception.

In receive antenna diversity; the receiver that has multiple antennas receives multiple
replicas of the same transmitted signal, assuming that the transmission came from the same
source. This holds true for SIMO channels. If the signal path between each antenna pair fades
independently, then when one path is in a fade, it is extremely unlikely that all the other paths
are also in deep fade. Therefore, the loss of signal power due to fade in one path is countered
by the same signal but received through a different path. This is like a line of soldiers. When
one soldier falls in battle, another is ready to take his place. Hence, extending this analogy
further, the more the soldiers, the stronger the line. The same is the argument in diversity.
The more the diversity, the easier can combat fades in a channel. Diversity is characterized
by the number of independent fading branches, or paths (routes). These paths are also known
as diversity order and are equal to the number of receive antennas in SIMO channels.
Logically, the higher the diversity order (independent fading paths, or receive antennas), the
better can combat fading. If the number of receive antennas tends to infinity, the diversity
order tends to infinity and the channel tends to additive white Gaussian noise (AWGN).

2.4.3 Spatial Multiplexing Gain

Spatial multiplexing offers a linear (in the number of transmit-receive antenna pairs or
min (Nr ,Nt) increase in the transmission rate (or capacity) for the same bandwidth and with

12
 An Overview of MIMO Wireless Systems

no additional power expenditure. It is only possible in MIMO channels [5, 30]. Consider the
case of two-transmit antennas and two-receive antennas. This can be extended to more
general MIMO channels.
The bit stream is split into two half-rate bit streams, modulated and transmitted
simultaneously from both the antennas. The receiver, having complete knowledge of the
channel, recovers these individual bit streams and combines them so as to recover the original
bit stream. Since the receiver has knowledge of the channel it provides receive diversity, but
the system has no transmit diversity since the bit streams are completely different from each
other in that they carry totally different data. Thus spatial multiplexing increases the
transmission rates proportionally with the number of transmit-receive antenna pairs.

2.5 Signal Model

In the transmission of digital information over a communication channel, the

modulator is the interface device that maps the digital information into analog waveforms
that match the characteristics of the channel. The mapping is generally performed by taking
blocks of k=log2M binary digits at a time from the information sequence and selecting one of
M=2k deterministic, finite, energy waveforms for transmission over the channel. When
analyzing communication systems, it is often unnecessary to model the up- and down-
conversion between the base-band models, or equivalent low pass signals and channel, which
then become complex valued. Throughout this thesis, complex baseband representation of
signals is used. In the sequel, QAM modulation is selected. The average energy per transmit
symbol equals to Es =(2/3) (M− 1) and the average energy per transmit bit equals to Eb= Es/k
[27, 28].

2.5.1 Basic MIMO Channel Model

The basic MIMO wireless channel is illustrated in Figure 2-3 [29]

13
 An Overview of MIMO Wireless Systems

Figure 2-3 MIMO Channel Model.

The Nt×1 input data vector is processed through a digital signal processor (DSP) and mapped
to the Nt×1 transmit symbol vector s which is forwarded to the transmit antenna array for
transmission across the MIMO wireless channel. Typically, the transmit symbol vector s is
complex with elements belonging to an M-ary digital modulation scheme such as QPSK, M-
QAM, etc. In a general sense, the transmitted symbols are perturbed during transmission by
the diffraction, scattering, and reflection characteristics of the wireless channel, as well as
noise inherent in the channel [31].

The MIMO channel itself is modeled as an Nr×Nt matrix H = [h1, h2,…,hnt]:

ℎ11 ℎ12 ⋯ ℎ1𝑁𝑡 1st receive antenna

ℎ ℎ22 ⋯ ℎ2𝑁𝑡
𝐻 = 21 (2.3)
⋮ ⋮ ⋱ ⋮
ℎ𝑁𝑟 1 ℎ𝑁𝑟 1 ⋯ ℎ𝑁𝑟 𝑁𝑡

2nd transmit antenna

The coefficient hij denotes the single-input single-output (SISO) channel fading between the
jth (j= 1,2, … ,Nt) transmit antenna and the ith (i=1,2,…,Nr) receive antenna. The column
vector hj=[h1j,h2j, … , hNrj]T is the single-input multiple-output (SIMO) channel produced by
the jth transmit antenna through the nr receive antennas. The row vector 𝐡𝑇𝑖 = [hi1,hi2, … , hiNt]
is the multiple-input single-output (SIMO) channel that represents the different Nt paths
arriving to the ith receive antenna.

14
 An Overview of MIMO Wireless Systems

Although not explicitly shown in Figure 2-3, the transmitted symbols are perturbed by
additive noise during transmission. In most environments, an Nr×1 channel noise vector n
with elements ni  CN(0,N0) is modeled as AWGN.
The receiver antenna array produces an Nr ×1 receive vector r, expressed efficiently
through vector-matrix notation as:
r = Hs + n (2.4)

It should be noted that Figure 2-3 is highly simplified to emphasize the MIMO
specific components of the narrowband wireless system in equivalent baseband. The input
vector s is generated by the application specific modules which would typically include the
source encoder, channel encoder, data interleaver, input multiplexer, etc. The digitally
modulated symbols in practice would also be up-converted to RF and amplified prior to
transmission. From Figure 2-3 and the MIMO composite channel matrix of (2.5), the
composite channel power gain between the ith receive and jth transmit antennas is given simply
as |hij|2. Similarly, the composite channel power gain, from the transmit array, as seen by the ith
receive antenna is given as the square of ||⋅||2 of the ith row of H [29]. This is illustrated in
Figure 2-4.

Figure 2-4 MIMO Channel to ith Receiver.

The composite channel power gain from the jth transmit antenna, as seen at the receive
antenna array is computed as the square of ||⋅||2 of the jth column of H. This is illustrated in
Figure 2-5,

15
 An Overview of MIMO Wireless Systems

Figure 2-5 MIMO Channel from j Transmitter.

th

The total composite channel power gain of the MIMO system is given as the square of ||⋅||F of
H or Tr(HHH).

2.6 MIMO Information Theory

Some performance limits of MIMO systems are given in this section in terms of
spectral efficiency with reliability. Let the random variables x and y be the input and the
output of a memoryless wireless channel. The observation of the channel output y gives us
information about the variable x. The mutual information I(x; y) is defined by the information
theory to measure the amount of information that y contains about x. The maximization of the
mutual information over all possible input distributions p(x) determines the maximum data
rate that a channel can support without error, also known as channel capacity. The channel
capacity is then measured in bits per channel use. Commonly, it is represented within a unit
bandwidth of the channel and it is measured in bits/s/Hz.

For a discrete memoryless channel, the channel capacity is defined by [29]

𝐶 𝑖𝑛𝑠𝑡 = max𝑝(𝑥) 𝐼(𝑥; 𝑦) (2.5)

When constraining the available power at the transmitter to PT, the channel capacity becomes

𝐶 𝑖𝑛𝑠𝑡 = max𝑝 𝑥 ,𝐸(𝑥 † 𝑥)≤𝑃𝑇 𝐼(𝑥; 𝑦) (2.6)

In [11] and [29], the capacity of MIMO channel was evaluated for different channel time
variations (ergodic and non-ergodic).

16
 An Overview of MIMO Wireless Systems

2.6.1 Ergodic Capacity

To show the gain offered by the use of multiple antennas at both sides in terms of
capacity, we begin with fundamental results derived for single antenna (SISO) or multiple
antennas at one side (SIMO or MISO) wireless systems.
In the following, the channel is ergodic and flat fading. Perfect CSI (Channel State
Information) is available only at the receiver and the transmitter is constrained in its total
power to PT, i.e. E(x†x) ≤ PT. The ergodic capacity is defined as the expectation of the
instantaneous channel capacity (2.5) over the distribution of the elements of the channel
matrix H

𝐶 𝑒𝑟𝑔 = 𝐸𝐻 max𝑝 𝑥 ,𝐸(𝑥 † 𝑥)≤𝑃𝑇 𝐼(𝑥; 𝑦) (2.7)

2.6.1.1 Capacity of SISO Channel

For a memoryless SISO system (Nt = Nr = 1), the channel matrix H is reduced to a
scalar complex variable h. The ergodic capacity is given by

𝑒𝑟𝑔
𝐶𝑆𝐼𝑆𝑂 = 𝐸ℎ max𝑝 𝑥 ,𝐸( 𝑥 2 )≤𝑃𝑇 𝐼(𝑥; 𝑦) (2.8)

For a SISO channel, the ergodic capacity (2.8) can be written as

𝑒𝑟𝑔
𝐶𝑆𝐼𝑆𝑂 = 𝐸ℎ 𝑙𝑜𝑔2 (1 + 𝜌 ℎ 2 ) (2.9)

where  denotes the average signal-to-noise ratio (SNR) per receive antenna.
If |h|2 = 1, the instantaneous capacity equals

𝑖𝑛𝑠𝑡
𝐶𝑆𝐼𝑆𝑂 = 𝑙𝑜𝑔2 (1 + 𝜌) (2.10)

It increases slowly with respect to the SNR, according to the logarithm of (1+). For high
SNR, it is noticed that a gain of 3 dB on  will provide only one bit increase in capacity.
When the channel gain amplitude |h| is Rayleigh distributed, |h|2 follows a chi-squared
distribution with two degrees of freedom which leads to an exponential distribution. Equation
(2.9) can then be written as

17
 An Overview of MIMO Wireless Systems

𝑒𝑟𝑔
𝐶𝑆𝐼𝑆𝑂 = 𝐸ℎ 𝑙𝑜𝑔2 (1 + 𝜌22 ) (2.11)

where 22 is a chi-square distributed random variable with two degrees of freedom.

2.6.1.2 Capacity of SIMO and MISO Channels

Consider a SIMO channel h = [h1,h2, . . . ,hNr ], with a single transmit and Nr receive
antennas. The capacity under ergodicity assumption is given by [29]

𝑒𝑟𝑔 𝑁𝑡
𝐶𝑆𝐼𝑀𝑂 = 𝐸ℎ 𝑙𝑜𝑔2 1 +  𝑖=1 ℎ𝑖 2
(2.12)

Like in the SISO case, if we assume that h satisfies |hi| = 1, i = 1, · · · ,Nt, then the
𝑖𝑛𝑠𝑡
instantaneous capacity becomes 𝐶𝑆𝐼𝑀𝑂 = 𝑙𝑜𝑔2 (1 + 𝜌𝑁𝑟 ) . Thus, the addition of receive
antennas only results in a logarithmic increase of the capacity with the SNR.
With optimal combining at the receiver, the capacity of a Rayleigh fading SIMO channel can
be expressed as

𝑒𝑟𝑔
𝐶𝑆𝐼𝑀𝑂 = 𝐸ℎ 𝑙𝑜𝑔2 (1 + 𝜌22𝑁 ) (2.13)
𝑟

where 22𝑁 is a chi-square distributed random variable with 2Nr degrees of freedom.
𝑟

When multiple antennas are employed only at the transmitter, the capacity of the MISO channel
is given by

𝑒𝑟𝑔 𝜌 𝑁𝑡 2
𝐶𝑀𝐼𝑆𝑂 = 𝐸ℎ 𝑙𝑜𝑔2 1 + 𝑁 𝑖=1 ℎ𝑖 (2.14)
𝑡

If |hi| = 1, i = 1, · · · ,Nt, then the instantaneous capacity is equal to

𝑖𝑛𝑠𝑡
𝐶𝑀𝐼𝑆𝑂 = 𝑙𝑜𝑔2 (1 + 𝜌) (2.15)

There is no gain in capacity over a SISO channel. By comparing equations (2.12) and (2.14)
and assuming the same total number of antennas, it is clear that CMISO is lower than CSIMO
when CSI is not available at the transmitter.

18
 An Overview of MIMO Wireless Systems

2.6.1.3 Capacity of MIMO Channel

The average capacity of a random ergodic MIMO channel is given by [29]

𝑒𝑟𝑔
𝐶𝑀𝐼𝑀𝑂 = 𝐸𝐻 max𝑝 𝑥 ;𝑡𝑟 (𝑄)≤𝑃𝑇 𝐼(𝑥; 𝑦) (2.16)

where Q = E(xx†) is the transmit signal covariance matrix, and tr(Q) = E(x†x) denotes the
trace of Q.
The mutual information is maximized for a zero mean circularly symmetric complex
Gaussian distributed input. The capacity is then given by

𝑒𝑟𝑔 1
𝐶𝑀𝐼𝑀𝑂 = 𝐸𝐻 𝑙𝑜𝑔2 𝑑𝑒𝑡 𝐼𝑁𝑟 + 2𝑁 𝐻𝑄𝐻 † (2.17)
0

When no CSI is available at the transmitter, the available power PT can be uniformly
distributed among the transmit antennas. For uncorrelated channel, the transmit covariance
𝑃𝑇
matrix is equal to 𝑃 = 𝐼 and the corresponding channel capacity becomes
𝑁𝑡 𝑁𝑡

𝑒𝑟𝑔 𝑃
𝐶𝑀𝐼𝑀𝑂 = 𝐸𝐻 𝑙𝑜𝑔2 𝑑𝑒𝑡 𝐼𝑁𝑟 + 2𝑁 𝑇𝑁 𝐻𝐻 † (2.18)
0 𝑡

𝑃
Let 𝜌 = 2𝑁𝑇 be the average SNR per receive antenna. For optimal combining between Nr
0

antennas at the receiver, the capacity can be written as

𝑒𝑟𝑔 𝜌
𝐶𝑀𝐼𝑀𝑂 = 𝐸𝐻 𝑁𝑡 𝑙𝑜𝑔2 1 + 𝑁 22𝑁 (2.19)
𝑡 𝑟

1
By the law of large numbers, the term 𝑁 𝐻𝐻 † INr as Nt gets larger and Nr remains fixed
𝑡

[29]. Thus, the ergodic capacity in this case is:

𝑒𝑟𝑔
𝐶𝑀𝐼𝑀𝑂 = 𝐸𝐻 𝑁𝑟 𝑙𝑜𝑔2 1 + 𝜌 (2.20)

Hence, the capacity reaches an asymptotic value for a fixed Nr. It is then un-advantageous to
increase indefinitely the number of transmit antennas.

19
 An Overview of MIMO Wireless Systems

Further analysis of the MIMO channel capacity could be conducted by applying the singular
value decomposition (SVD) to the channel matrix H, that is

H = UDV † (2.21)

The Nr × Nr and Nt × Nt complex matrices U and V are unitary and the Nr × Nt non-negative
diagonal matrix D contains the singular values of the matrix H. Substituting H by its
decomposition in (1.14) leads to the following capacity expression:

𝑚
𝑒𝑟𝑔 𝜌
𝐶𝑀𝐼𝑀𝑂 = 𝐸𝐻 𝑙𝑜𝑔2 1 + 
𝑁𝑡 𝑖
𝑖=1

𝑚 𝜌
= 𝑖=1 𝐸𝑖 𝑙𝑜𝑔2 1 + 𝑁 𝑖 (2.22)
𝑡

where ( 𝑖 ) 1≤𝑖≤𝑚 are the square of the non-zero entries of the diagonal matrix

𝐷 = 𝑑𝑖𝑎𝑔 1 , 2 , … , 𝑚 , 0, … ,0 (2.23)

and m = min(Nt , Nr) is equal to the channel rank. We can deduce that the total capacity of a
given MIMO channel H is made up by the sum of m parallel independent AWGN SISO sub-
channels, whose channel gain equals respectively  𝑖 , 𝑖 = 1,2, … , 𝑚 . Consequently, the
MIMO capacity grows linearly with m = min(Nt , Nr) rather than logarithmically.

For slow-varying or block fading channel, the ergodicity property is not respected and the
classical capacity definition is no longer appropriate. Capacity becomes a random variable
which depends on the channel instantaneous response.
Figure 2-6 shows the ergodic capacity for different antenna configurations with Nt = Nr = 1,
2, 3, 4. It shows that as the number of antenna increases, the capacity increases for a given
SNR.

20
 An Overview of MIMO Wireless Systems

25
M=2
M=3
M=4 Nt=Nr=4
20
M=1
Nt=Nr=3
Ergodic Capacity (Bits/sec/Hz)

15
Nt=Nr=2

Nt=Nr=1
10

0
0 2 4 6 8 10 12 14 16 18 20
SNR (dB)

Figure 2-6 Ergodic capacity for different antenna configurations with Nt = Nr

2.6.2 Outage Capacity

In reality, the block lengths are finite. The common example is speech transmission.
In such cases, outage capacity is considered. Outage capacity is the capacity that is
guaranteed with a certain level of reliability. p% outage capacity is defined as the information
rate that is guaranteed for (100-p)% of the channel realizations. Figure 2-7 shows the 10%
outage capacity for several MIMO cases, when the channel is i.i.d. and unknown at the
transmitter. We note that as the SNR increases, the capacity increases and as the number of
antennas increases, so does the capacity.

21
 An Overview of MIMO Wireless Systems

20
M=1
18 M=2
M=3
16 M=4 Nt=Nr=4
10% Outage Capacity (Bits/sec/Hz)

14 Nt=Nr=3
12
Nt=Nr=2
10

8
Nt=Nr=1
6

0
0 2 4 6 8 10 12 14 16 18 20
SNR (dB)

Figure 2-7 10% outage capacity for various antenna configurations.

22