04-03-2013, 04:20 PM
AN OVERVIEW OF SMART ANTENNA TECHNOLOGY FOR HETEROGENEOUS NETWORKS
[attachment=52558]
ABSTRACT
This article presents an overview of advances in the smart antenna technologies that are
currently considered as a potential candidate for interference cancellation in wireless
communications. The motivation is to assimilate the advances in software radio with the
optimized space-time interference rejection schemes. The solutions currently envisioned for
use in various networks need to be visualized to map them into re-configurable architectures
and realize network support for heterogeneity. The conclusion broadly drawn is that the
advancement in DSP and ASIC technologies will make the major contribution in making
these technologies technically and commercially realistic.
INTRODUCTION
Advances in new mobile applications coupled with the
complexity of radio communication systems have
paved the way to new f lexible algorithms at the
receiver and transmitter in communications systems.
One of the most popular spatio-temporal interference cancellation
techniques is realized by applying the smart antenna
concept. In the prevailing communication landscapes, as shown
in Fig. 1, networks are heterogeneous and this imminently dictates
the smart antenna to have the capability to operate with
not only Univer sal Mobile Telecommunication Systems
(UMTS) standards but also with other multiple radio standards
in third-generation mobile networks. The drive for a better
grade of service and the need for serving the exponentially
growing number of users demand greater synergy between
multiple access systems and the spatial filtering technology.
Fading resistant beamforming requires better understanding of
channel models as well as the development of robust adaptive
algorithms. Exploiting the scientific and technology base in
space-time receiver modeling is of the utmost importance. This
will set the scene to better identify the components that will
support advanced base stations, eventually resulting in technology-
translators via tailor-made solutions for increased capacity.
HETEROGENEOUS NETWORKING WITH
SPACE-TIME INTERFERENCE CANCELLERS
In recent years there has been rapid progress in telecommunications,
resulting in new application scenarios for mobile
networks (Fig. 1). With advances in technology, a variety of
communication devices such as Ethernet, WaveLAN, CDPD,
Metricom Ricochet, and cellular modems have become available
at affordable prices. Today, it is common for a laptop to
have access to more than one network. All these technologies
offer different network characteristics, leading to heterogeneity
in network architectures. To deal with heterogeneous networks,
important components of space-time processors need
to be identified with a goal to make information services and
applications ubiquitous and flexibly available to people on the
move.
GENESIS OF SPATIO-TEMPORAL
INTERFERENCE CANCELLATION IN
TDMA AND CDMA NETWORKS
There are many ways in which spatio-temporal filtering can be
used in a mobile system allowing performance to be traded
against the complexity of implementation. Indeed, in some
cases simpler techniques give better performance because they
are more robust in difficult propagation environments.
Beamsteering refers to the class of algorithms that attempt
to direct a beam toward the wanted mobile but make no
attempt to null co-channel interference signals. The interference
reduction effectiveness of the basic beamsteering may be
enhanced by combining it with other techniques such as frequency
hopping.
Adaptive nulling can be used to further reduce co-channel
interference in the uplink to improve the overall capacity of
the system. In TDMA systems there are two principal ways in
which adaptive nulling can be used:
• Spatial filtering for interference reduction (SFIR) — In
this scheme nulls are formed in the direction of interference
sources in uplink and downlink. This improves the
carrier to interference (C/I) ratio and allows the frequency
reuse pattern to be tightened, thus increasing capacity.
• Spatial division multiple access (SDMA) — This involves
the use of adaptive nulling to allow two or more mobiles
in the same cell to share the same frequency and time
slot. One beam is formed for each mobile with nulls in
the direction of the other mobiles.
RELATIONSHIP WITH THE DUPLEXING SCHEMES
In the traditional cellular systems, only the uplink channel can
be estimated by the base station and therefore the downlink
beam pattern must somehow be derived from estimates of the
uplink channel. The perceived direction of arrival (DOA) of a
signal (or, more precisely, its source vector) is a function of
both time and frequency in a time varying dispersive channel.
The air interface duplex scheme is therefore s ignificant
because it affects the degree of correlation that exists between
the uplink and downlink channels.
The performance improvement obtainable from the smart
antenna system is not symmetrical between uplink and downlink
for this reason. The degree of asymmetry depends on the
class of algorithm being employed. The direction of a beam
maximum is less sensitive to time and frequency translation
than the direction of a null. Therefore, the asymmetry is particularly
evident when it is necessary to form deep nulls in the
downlink beam pattern, which is a requirement for SDMA.
In TDD schemes the uplink and downlink channels may
be considered reciprocal if the source vectors of the signals
do not change significantly in the time between the transmit
and receive time slots. The rate of change of the source vector
is controlled by the mobile velocity and the angular
spread of the channel. In a macrocell environment where
the angular spreading is low the source vector of the mobile
may not change significantly over a distance of many wavelengths.
In this situation the source vector is not significantly
altered by the fast fading. However, in microcell
environments, moving only a small fraction of a wavelength
can be enough to significantly change the source vector of
the mobile and therefore move out of a null. Forming deep
downlink nul ls in microcel l environments is therefore
impractical in almost all cases. An appropriate choice in
many systems may be to employ SFIR on the uplink and
beamsteering for the downlink.
[attachment=52558]
ABSTRACT
This article presents an overview of advances in the smart antenna technologies that are
currently considered as a potential candidate for interference cancellation in wireless
communications. The motivation is to assimilate the advances in software radio with the
optimized space-time interference rejection schemes. The solutions currently envisioned for
use in various networks need to be visualized to map them into re-configurable architectures
and realize network support for heterogeneity. The conclusion broadly drawn is that the
advancement in DSP and ASIC technologies will make the major contribution in making
these technologies technically and commercially realistic.
INTRODUCTION
Advances in new mobile applications coupled with the
complexity of radio communication systems have
paved the way to new f lexible algorithms at the
receiver and transmitter in communications systems.
One of the most popular spatio-temporal interference cancellation
techniques is realized by applying the smart antenna
concept. In the prevailing communication landscapes, as shown
in Fig. 1, networks are heterogeneous and this imminently dictates
the smart antenna to have the capability to operate with
not only Univer sal Mobile Telecommunication Systems
(UMTS) standards but also with other multiple radio standards
in third-generation mobile networks. The drive for a better
grade of service and the need for serving the exponentially
growing number of users demand greater synergy between
multiple access systems and the spatial filtering technology.
Fading resistant beamforming requires better understanding of
channel models as well as the development of robust adaptive
algorithms. Exploiting the scientific and technology base in
space-time receiver modeling is of the utmost importance. This
will set the scene to better identify the components that will
support advanced base stations, eventually resulting in technology-
translators via tailor-made solutions for increased capacity.
HETEROGENEOUS NETWORKING WITH
SPACE-TIME INTERFERENCE CANCELLERS
In recent years there has been rapid progress in telecommunications,
resulting in new application scenarios for mobile
networks (Fig. 1). With advances in technology, a variety of
communication devices such as Ethernet, WaveLAN, CDPD,
Metricom Ricochet, and cellular modems have become available
at affordable prices. Today, it is common for a laptop to
have access to more than one network. All these technologies
offer different network characteristics, leading to heterogeneity
in network architectures. To deal with heterogeneous networks,
important components of space-time processors need
to be identified with a goal to make information services and
applications ubiquitous and flexibly available to people on the
move.
GENESIS OF SPATIO-TEMPORAL
INTERFERENCE CANCELLATION IN
TDMA AND CDMA NETWORKS
There are many ways in which spatio-temporal filtering can be
used in a mobile system allowing performance to be traded
against the complexity of implementation. Indeed, in some
cases simpler techniques give better performance because they
are more robust in difficult propagation environments.
Beamsteering refers to the class of algorithms that attempt
to direct a beam toward the wanted mobile but make no
attempt to null co-channel interference signals. The interference
reduction effectiveness of the basic beamsteering may be
enhanced by combining it with other techniques such as frequency
hopping.
Adaptive nulling can be used to further reduce co-channel
interference in the uplink to improve the overall capacity of
the system. In TDMA systems there are two principal ways in
which adaptive nulling can be used:
• Spatial filtering for interference reduction (SFIR) — In
this scheme nulls are formed in the direction of interference
sources in uplink and downlink. This improves the
carrier to interference (C/I) ratio and allows the frequency
reuse pattern to be tightened, thus increasing capacity.
• Spatial division multiple access (SDMA) — This involves
the use of adaptive nulling to allow two or more mobiles
in the same cell to share the same frequency and time
slot. One beam is formed for each mobile with nulls in
the direction of the other mobiles.
RELATIONSHIP WITH THE DUPLEXING SCHEMES
In the traditional cellular systems, only the uplink channel can
be estimated by the base station and therefore the downlink
beam pattern must somehow be derived from estimates of the
uplink channel. The perceived direction of arrival (DOA) of a
signal (or, more precisely, its source vector) is a function of
both time and frequency in a time varying dispersive channel.
The air interface duplex scheme is therefore s ignificant
because it affects the degree of correlation that exists between
the uplink and downlink channels.
The performance improvement obtainable from the smart
antenna system is not symmetrical between uplink and downlink
for this reason. The degree of asymmetry depends on the
class of algorithm being employed. The direction of a beam
maximum is less sensitive to time and frequency translation
than the direction of a null. Therefore, the asymmetry is particularly
evident when it is necessary to form deep nulls in the
downlink beam pattern, which is a requirement for SDMA.
In TDD schemes the uplink and downlink channels may
be considered reciprocal if the source vectors of the signals
do not change significantly in the time between the transmit
and receive time slots. The rate of change of the source vector
is controlled by the mobile velocity and the angular
spread of the channel. In a macrocell environment where
the angular spreading is low the source vector of the mobile
may not change significantly over a distance of many wavelengths.
In this situation the source vector is not significantly
altered by the fast fading. However, in microcell
environments, moving only a small fraction of a wavelength
can be enough to significantly change the source vector of
the mobile and therefore move out of a null. Forming deep
downlink nul ls in microcel l environments is therefore
impractical in almost all cases. An appropriate choice in
many systems may be to employ SFIR on the uplink and
beamsteering for the downlink.