U.S. patent application number 13/937736 was filed with the patent office on 2013-11-07 for distributed downlink coordinated multi-point (comp) framework.
The applicant listed for this patent is QUALCOMM Incorporated. Invention is credited to Alan Barbieri, Naga Bhushan, Alexei Y. Gorokhov, Siddhartha Mallik.
Application Number | 20130294275 13/937736 |
Document ID | / |
Family ID | 41722277 |
Filed Date | 2013-11-07 |
United States Patent
Application |
20130294275 |
Kind Code |
A1 |
Gorokhov; Alexei Y. ; et
al. |
November 7, 2013 |
DISTRIBUTED DOWNLINK COORDINATED MULTI-POINT (CoMP) FRAMEWORK
Abstract
Systems and methodologies are described that facilitate
dynamically forming clusters in a wireless communication
environment. A set of non-overlapping clusters can be formed
dynamically over time and in a distributed manner. Each of the
clusters can include a set of base stations and a set of mobile
devices. The clusters can be yielded based upon a set of local
strategies selected by base stations across the network converged
upon through message passing. For example, each base station can
select a particular local strategy as a function of time based upon
network-wide utility estimates respectively conditioned upon
implementation of the particular local strategy and disparate
possible local strategies that can cover the corresponding base
station. Moreover, operation within each of the clusters can be
coordinated.
Inventors: |
Gorokhov; Alexei Y.; (San
Diego, CA) ; Mallik; Siddhartha; (San Diego, CA)
; Bhushan; Naga; (San Diego, CA) ; Barbieri;
Alan; (San Diego, CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
QUALCOMM Incorporated |
San Diego |
CA |
US |
|
|
Family ID: |
41722277 |
Appl. No.: |
13/937736 |
Filed: |
July 9, 2013 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12547395 |
Aug 25, 2009 |
8498647 |
|
|
13937736 |
|
|
|
|
61092490 |
Aug 28, 2008 |
|
|
|
Current U.S.
Class: |
370/252 ;
370/328 |
Current CPC
Class: |
H04W 24/02 20130101;
H04B 17/382 20150115; H04B 7/024 20130101 |
Class at
Publication: |
370/252 ;
370/328 |
International
Class: |
H04W 24/02 20060101
H04W024/02 |
Claims
1. A method of wireless communication, comprising: choosing, by a
base station, a particular local strategy as a function of time
based upon network-wide utility estimates respectively conditioned
upon the particular local strategy and disparate possible local
strategies; and controlling, by the base station, operation within
a cluster dynamically formed based upon the chosen particular local
strategy.
2. The method of claim 1, further comprising exchanging information
utilized to evaluate the network-wide utility estimates with at
least one neighbor base station.
3. The method of claim 2, wherein the information comprises a
cooperative utility value that reflects an estimate of total
utility assuming cooperation between a source and a target and a
non-cooperative utility value, and a non-cooperative utility value
that reflects an estimate of total utility assuming lack of
cooperation between the source and the target.
4. The method of claim 2, wherein the information comprises a
plurality of utility values assuming various constraints upon a
target, wherein the assumed constraints are reported from a source
to the target.
5. The method of claim 1, wherein the particular local strategy and
the disparate possible local strategies each cover one or more base
stations, one or more mobile devices served by the one or more base
stations, and underlying antenna weights and power spectral
densities for the one or more base stations to serve the one or
more mobile devices.
6. The method of claim 1, wherein the particular local strategy and
the disparate possible local strategies each are subject to a
limited maximum order constraint.
7. The method of claim 1, wherein the cluster and disparate
clusters dynamically formed in a network are non-overlapping.
8. The method of claim 1, wherein transmission information is
exchanged between the cluster and disparate clusters dynamically
formed in the network to enable assessing inter-cluster
interference.
9. An apparatus, comprising: means for choosing, by a base station,
a particular local strategy as a function of time based upon
network-wide utility estimates respectively conditioned upon the
particular local strategy and disparate possible local strategies;
and means for controlling, by the base station, operation within a
cluster dynamically formed based upon the chosen particular local
strategy.
10. The apparatus of claim 9, further comprising means for
exchanging information utilized to evaluate the network-wide
utility estimates with at least one neighbor base station.
11. The apparatus of claim 10, wherein the information comprises a
cooperative utility value that reflects an estimate of total
utility assuming cooperation between a source and a target and a
non-cooperative utility value, and a non-cooperative utility value
that reflects an estimate of total utility assuming lack of
cooperation between the source and the target.
12. The apparatus of claim 10, wherein the information comprises a
plurality of utility values assuming various constraints upon a
target, wherein the assumed constraints are reported from a source
to the target.
13. The apparatus of claim 9 wherein the particular local strategy
and the disparate possible local strategies each cover one or more
base stations, one or more mobile devices served by the one or more
base stations, and underlying antenna weights and power spectral
densities for the one or more base stations to serve the one or
more mobile devices.
14. The apparatus of claim 9, wherein the particular local strategy
and the disparate possible local strategies each are subject to a
limited maximum order constraint.
15. The apparatus of claim 9, wherein the cluster and disparate
clusters dynamically formed in a network are non-overlapping.
16. The apparatus of claim 9, wherein transmission information is
exchanged between the cluster and disparate clusters dynamically
formed in the network to enable assessing inter-cluster
interference.
17. A computer program product, comprising: a computer-readable
medium comprising: code for causing at least one computer to select
a particular local strategy that includes a base station as a
function of time based upon network-wide utility estimates
respectively conditioned upon implementation of the particular
local strategy and disparate possible local strategies that include
the base station; and code for causing at least one computer to
coordinate operation within a cluster formed according to the
selected particular local strategy.
18. The computer program product of claim 17, wherein the
computer-readable medium further comprises code for causing at
least one computer to evaluate local utilities of the particular
local strategy and the disparate possible local strategies.
19. The computer program product of claim 18, wherein the
computer-readable medium further comprises code for causing at
least one computer to exchange strategy and utility information
with one or more neighbor base stations via iterative message
passing.
20. The computer program product of claim 17, wherein the
computer-readable medium further comprises code for causing at
least one computer to yield the network-wide utility estimates.
21. The computer program product of claim 17, wherein the
particular local strategy and the disparate possible local
strategies each cover one or more base stations, one or more mobile
devices served by the one or more base stations, and underlying
antenna weights and power spectral densities for the one or more
base stations to serve the one or more mobile devices.
22. The computer program product of claim 17, wherein the
particular local strategy and the disparate possible local
strategies each are subject to a limited maximum order
constraint.
23. The computer program product of claim 17, wherein the cluster
and disparate clusters dynamically formed in a network are
non-overlapping.
24. The computer program product of claim 17, wherein transmission
information related to one or more of beams or power spectral
densities (PSDs) is exchanged between the cluster and disparate
clusters dynamically formed in the network to enable assessing
inter-cluster interference.
25. An apparatus, comprising: a clustering component that
dynamically selects a local strategy to implement with a base
station from a set of possible local strategies, wherein the
possible local strategies enable the base station to cooperate with
one or more neighbor base stations; a metric evaluate component
that analyzes local utilities of the possible local strategies in
the set; and a negotiation component that employs message passing
to agree on compatible local strategies across a network.
26. The apparatus of claim 25, further comprising a cooperation
component that coordinates operation of the base station and one or
more cooperating base stations included in a common cluster formed
based upon the selected local strategy.
Description
RELATED APPLICATIONS
[0001] The present application for patent is a continuation of U.S.
patent application Ser. No. 12/547,395, entitled, "DISTRIBUTED
DOWNLINK COORDINATED MULTI-POINT (CoMP) FRAMEWORK," filed on Aug.
25, 2009, which claims priority to Provisional Application No.
61/092,490 entitled "DISTRIBUTED DL COOPERATION FRAMEWORK FOR USE
IN MIMO SYSTEMS" filed Aug. 28, 2008, and assigned to the assignee
hereof and hereby expressly incorporated by reference herein.
BACKGROUND
[0002] 1. Field
[0003] The following description relates generally to wireless
communications, and more particularly to dynamically selecting
clustering strategies in a distributed manner in a wireless
communication environment that employs downlink coordinated
multi-point (CoMP).
[0004] 2. Background
[0005] Wireless communication systems are widely deployed to
provide various types of communication content such as, for
example, voice, data, and so on. Typical wireless communication
systems can be multiple-access systems capable of supporting
communication with multiple users by sharing available system
resources (e.g., bandwidth, transmit power, . . . ). Examples of
such multiple-access systems can include code division multiple
access (CDMA) systems, time division multiple access (TDMA)
systems, frequency division multiple access (FDMA) systems,
orthogonal frequency division multiple access (OFDMA) systems, and
the like. Additionally, the systems can conform to specifications
such as third generation partnership project (3GPP), 3GPP long term
evolution (LTE), ultra mobile broadband (UMB), and/or multi-carrier
wireless specifications such as evolution data optimized (EV-DO),
one or more revisions thereof, etc.
[0006] Generally, wireless multiple-access communication systems
can simultaneously support communication for multiple mobile
devices. Each mobile device can communicate with one or more base
stations via transmissions on forward and reverse links. The
forward link (or downlink) refers to the communication link from
base stations to mobile devices, and the reverse link (or uplink)
refers to the communication link from mobile devices to base
stations. Further, communications between mobile devices and base
stations can be established via single-input single-output (SISO)
systems, multiple-input single-output (MISO) systems,
multiple-input multiple-output (MIMO) systems, and so forth. In
addition, mobile devices can communicate with other mobile devices
(and/or base stations with other base stations) in peer-to-peer
wireless network configurations.
[0007] Traditionally, in a wireless communication network with
multiple base stations and multiple mobile devices, each mobile
device is typically associated with a particular one of the
multiple base stations. For instance, a mobile device can be
associated with a given base station as a function of various
factors such as signal strength, Channel Quality Indicator (CQI),
and so forth. Thus, the mobile device can be served by the given
base station (e.g., uplink and downlink transmissions can be
exchanged there between, . . . ), while other base stations in
vicinity can generate interference.
[0008] Moreover, cooperation between base stations has become more
commonly leveraged. In particular, multiple base stations in a
wireless communication network can be interconnected, which can
allow for sharing data between base stations, communicating there
between, and so forth. For instance, in a wireless communication
network deployment across a city, base stations included in the
deployment can serve a set of mobile devices located within
proximity of the base stations. Such deployment oftentimes utilize
a common, centralized scheduler; thus, a scheduler decision can be
rendered to transmit from the base stations in the deployment to a
first mobile device during a first time period, a second mobile
device during a second time period, and so forth. However,
centralized scheduling can be difficult at best to perform.
Moreover, involvement of all (or most) base stations from the
deployment when serving a particular mobile device can be
impractical and unneeded due to connectivity between base
stations.
SUMMARY
[0009] The following presents a simplified summary of one or more
aspects in order to provide a basic understanding of such aspects.
This summary is not an extensive overview of all contemplated
aspects, and is intended to neither identify key or critical
elements of all aspects nor delineate the scope of any or all
aspects. Its sole purpose is to present some concepts of one or
more aspects in a simplified form as a prelude to the more detailed
description that is presented later.
[0010] In accordance with one or more embodiments and corresponding
disclosure thereof, various aspects are described in connection
with dynamically forming clusters in a wireless communication
environment. A set of non-overlapping clusters can be formed
dynamically over time and in a distributed manner. Each of the
clusters can include a set of base stations and a set of mobile
devices. The clusters can be yielded based upon a set of local
strategies selected by base stations across the network converged
upon through message passing. For example, each base station can
select a particular local strategy as a function of time based upon
network-wide utility estimates respectively conditioned upon
implementation of the particular local strategy and disparate
possible local strategies that can cover the corresponding base
station. Moreover, operation within each of the clusters can be
coordinated.
[0011] According to related aspects, a method is described herein.
The method can include evaluating local utilities of possible local
strategies involving a base station at a given time. Further, the
method can include exchanging strategy and utility information with
at least one neighbor base station through message passing.
Moreover, the method can include generating network-wide utility
estimates for the possible local strategies as a function of the
strategy and utility information received from the at least one
neighbor base station through message passing and the evaluated
local utilities. The method can also include selecting a particular
local strategy from the possible local strategies for use by the
base station based upon the network-wide utility estimates.
[0012] Another aspect relates to a wireless communications
apparatus. The wireless communications apparatus can include at
least one processor. The at least one processor can be configured
to analyze local utilities of possible local strategies. The at
least one processor can additionally be configured to implement
message passing to exchange strategy and utility information with
at least one neighbor base station. Moreover, the at least one
processor can be configured to estimate network-wide utilities for
the possible local strategies as a function of the strategy and
utility information obtained from the at least one neighbor base
station and the analyzed local utilities. Further, the at least one
processor can be configured to form a cluster based upon a
particular local strategy chosen from the possible local strategies
based upon the estimates of the network-wide utilities.
[0013] Yet another aspect relates to a wireless communications
apparatus. The wireless communications apparatus can include means
for choosing a particular local strategy as a function of time
based upon network-wide utility estimates respectively conditioned
upon the particular local strategy and disparate possible local
strategies. Moreover, the wireless communications apparatus can
include means for controlling operation within a cluster
dynamically formed based upon the chosen particular local
strategy.
[0014] Still another aspect relates to a computer program product
that can comprise a computer-readable medium. The computer-readable
medium can include code for causing at least one computer to select
a particular local strategy that includes a base station as a
function of time based upon network-wide utility estimates
respectively conditioned upon implementation of the particular
local strategy and disparate possible local strategies that include
the base station. Further, the computer-readable medium can include
code for causing at least one computer to coordinate operation
within a cluster formed according to the selected particular local
strategy.
[0015] Yet another aspect relates to an apparatus that can include
a clustering component that dynamically selects a local strategy to
implement with a base station from a set of possible local
strategies, wherein the possible local strategies enable the base
station to cooperate with one or more neighbor base stations.
Moreover, the apparatus can include a metric evaluate component
that analyzes local utilities of the possible local strategies in
the set. Further, the apparatus can include a negotiation component
that employs message passing to agree on compatible local
strategies across a network.
[0016] To the accomplishment of the foregoing and related ends, the
one or more aspects comprise the features hereinafter fully
described and particularly pointed out in the claims. The following
description and the annexed drawings set forth in detail certain
illustrative features of the one or more aspects. These features
are indicative, however, of but a few of the various ways in which
the principles of various aspects may be employed, and this
description is intended to include all such aspects and their
equivalents.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] FIG. 1 is an illustration of a wireless communication system
in accordance with various aspects set forth herein.
[0018] FIG. 2 is an illustration of an example system that
leverages a downlink cooperation framework that employs a
network-wide strategy where a set of base stations in a deployment
cooperatively operate.
[0019] FIG. 3 is an illustration of an example system that employs
dynamic clustering based upon a finite order strategy constraint in
a wireless communication environment.
[0020] FIG. 4 is an illustration of an example system that employs
distributed strategy negotiation in a wireless communication
environment.
[0021] FIG. 5 is an illustration of an example system that employs
message passing in a wireless communication environment.
[0022] FIG. 6 is an illustration of another example system that
employs message passing in a wireless communication
environment.
[0023] FIG. 7 is an illustration of an example system that supports
cooperation within clusters in a wireless communication
environment.
[0024] FIG. 8 is an illustration of an example system that employs
inter-site packet sharing (ISPS) (e.g., coherent ISPS, . . . )
within a cluster in a wireless communication environment.
[0025] FIG. 9 is an illustration of an example system that
implements cooperative beamforming within a cluster in a wireless
communication environment.
[0026] FIG. 10 is an illustration of an example system that
effectuates cooperative silence (CS) within a cluster in a wireless
communication environment.
[0027] FIG. 11 is an illustration of an example system in which
non-cooperative transmissions can be effectuated in a wireless
communication environment.
[0028] FIG. 12 is an illustration of an example system that
exchanges interference information as part of a message passing
strategy to manage non-cooperative interference in a wireless
communication environment.
[0029] FIGS. 13-15 illustrate example graphs associated with a
belief propagation framework for interference avoidance and CoMP
that can be implemented in connection with the techniques described
herein.
[0030] FIG. 16 is an illustration of an example methodology that
facilitates dynamically forming clusters in a wireless
communication environment.
[0031] FIG. 17 is an illustration of an example methodology that
facilitates leveraging cooperation between base stations in a
wireless communication environment.
[0032] FIG. 18 is an illustration of an example mobile device that
can be employed in connection with various aspects described
herein.
[0033] FIG. 19 is an illustration of an example system that
dynamically selects a local strategy to employ over time in a
wireless communication environment.
[0034] FIG. 20 is an illustration of an example wireless network
environment that can be employed in conjunction with the various
systems and methods described herein.
[0035] FIG. 21 is an illustration of an example system that enables
employing dynamically defined clusters in a wireless communication
environment.
DETAILED DESCRIPTION
[0036] Various aspects are now described with reference to the
drawings. In the following description, for purposes of
explanation, numerous specific details are set forth in order to
provide a thorough understanding of one or more aspects. It may be
evident, however, that such aspect(s) may be practiced without
these specific details.
[0037] As used in this application, the terms "component,"
"module," "system" and the like are intended to include a
computer-related entity, such as but not limited to hardware,
firmware, a combination of hardware and software, software, or
software in execution. For example, a component can be, but is not
limited to being, a process running on a processor, a processor, an
object, an executable, a thread of execution, a program, and/or a
computer. By way of illustration, both an application running on a
computing device and the computing device can be a component. One
or more components can reside within a process and/or thread of
execution and a component can be localized on one computer and/or
distributed between two or more computers. In addition, these
components can execute from various computer readable media having
various data structures stored thereon. The components can
communicate by way of local and/or remote processes such as in
accordance with a signal having one or more data packets, such as
data from one component interacting with another component in a
local system, distributed system, and/or across a network such as
the Internet with other systems by way of the signal.
[0038] Furthermore, various aspects are described herein in
connection with a terminal, which can be a wired terminal or a
wireless terminal. A terminal can also be called a system, device,
subscriber unit, subscriber station, mobile station, mobile, mobile
device, remote station, remote terminal, access terminal, user
terminal, terminal, communication device, user agent, user device,
or user equipment (UE). A wireless terminal can be a cellular
telephone, a satellite phone, a cordless telephone, a Session
Initiation Protocol (SIP) phone, a wireless local loop (WLL)
station, a personal digital assistant (PDA), a handheld device
having wireless connection capability, a computing device, or other
processing devices connected to a wireless modem. Moreover, various
aspects are described herein in connection with a base station. A
base station can be utilized for communicating with wireless
terminal(s) and can also be referred to as an access point, a Node
B, an Evolved Node B (eNode B, eNB), or some other terminology.
[0039] Moreover, the term "or" is intended to mean an inclusive
"or" rather than an exclusive "or." That is, unless specified
otherwise, or clear from the context, the phrase "X employs A or B"
is intended to mean any of the natural inclusive permutations. That
is, the phrase "X employs A or B" is satisfied by any of the
following instances: X employs A; X employs B; or X employs both A
and B. In addition, the articles "a" and "an" as used in this
application and the appended claims should generally be construed
to mean "one or more" unless specified otherwise or clear from the
context to be directed to a singular form.
[0040] The techniques described herein can be used for various
wireless communication systems such as code division multiple
access (CDMA), time division multiple access (TDMA), frequency
division multiple access (FDMA), orthogonal frequency division
multiple access (OFDMA), single carrier-frequency division multiple
access (SC-FDMA) and other systems. The terms "system" and
"network" are often used interchangeably. A CDMA system can
implement a radio technology such as Universal Terrestrial Radio
Access (UTRA), CDMA2000, etc. UTRA includes Wideband-CDMA (W-CDMA)
and other variants of CDMA. Further, CDMA2000 covers IS-2000, IS-95
and IS-856 standards. A TDMA system can implement a radio
technology such as Global System for Mobile Communications (GSM).
An OFDMA system can implement a radio technology such as Evolved
UTRA (E-UTRA), Ultra Mobile Broadband (UMB), IEEE 802.11 (Wi-Fi),
IEEE 802.16 (WiMAX), IEEE 802.20, Flash-OFDM, etc. UTRA and E-UTRA
are part of Universal Mobile Telecommunication System (UMTS). 3GPP
Long Term Evolution (LTE) is a release of UMTS that uses E-UTRA,
which employs OFDMA on the downlink and SC-FDMA on the uplink.
UTRA, E-UTRA, UMTS, LTE and GSM are described in documents from an
organization named "3rd Generation Partnership Project" (3GPP).
Additionally, CDMA2000 and Ultra Mobile Broadband (UMB) are
described in documents from an organization named "3rd Generation
Partnership Project 2" (3GPP2). Further, such wireless
communication systems can additionally include peer-to-peer (e.g.,
mobile-to-mobile) ad hoc network systems often using unpaired
unlicensed spectrums, 802.xx wireless LAN, BLUETOOTH and any other
short- or long-range, wireless communication techniques.
[0041] Single carrier frequency division multiple access (SC-FDMA)
utilizes single carrier modulation and frequency domain
equalization. SC-FDMA has similar performance and essentially the
same overall complexity as those of an OFDMA system. A SC-FDMA
signal has lower peak-to-average power ratio (PAPR) because of its
inherent single carrier structure. SC-FDMA can be used, for
instance, in uplink communications where lower PAPR greatly
benefits access terminals in terms of transmit power efficiency.
Accordingly, SC-FDMA can be implemented as an uplink multiple
access scheme in 3GPP Long Term Evolution (LTE) or Evolved
UTRA.
[0042] Various aspects or features described herein can be
implemented as a method, apparatus, or article of manufacture using
standard programming and/or engineering techniques. The term
"article of manufacture" as used herein is intended to encompass a
computer program accessible from any computer-readable device,
carrier, or media. For example, computer-readable media can include
but are not limited to magnetic storage devices (e.g., hard disk,
floppy disk, magnetic strips, etc.), optical disks (e.g., compact
disk (CD), digital versatile disk (DVD), etc.), smart cards, and
flash memory devices (e.g., EPROM, card, stick, key drive, etc.).
Additionally, various storage media described herein can represent
one or more devices and/or other machine-readable media for storing
information. The term "machine-readable medium" can include,
without being limited to, wireless channels and various other media
capable of storing, containing, and/or carrying instruction(s)
and/or data.
[0043] Referring now to FIG. 1, a wireless communication system 100
is illustrated in accordance with various embodiments presented
herein. System 100 comprises a base station 102 that can include
multiple antenna groups. For example, one antenna group can include
antennas 104 and 106, another group can comprise antennas 108 and
110, and an additional group can include antennas 112 and 114. Two
antennas are illustrated for each antenna group; however, more or
fewer antennas can be utilized for each group. Base station 102 can
additionally include a transmitter chain and a receiver chain, each
of which can in turn comprise a plurality of components associated
with signal transmission and reception (e.g., processors,
modulators, multiplexers, demodulators, demultiplexers, antennas,
etc.), as will be appreciated by one skilled in the art.
[0044] Base station 102 can communicate with one or more mobile
devices such as mobile device 116 and mobile device 122; however,
it is to be appreciated that base station 102 can communicate with
substantially any number of mobile devices similar to mobile
devices 116 and 122. Mobile devices 116 and 122 can be, for
example, cellular phones, smart phones, laptops, handheld
communication devices, handheld computing devices, satellite
radios, global positioning systems, PDAs, and/or any other suitable
device for communicating over wireless communication system 100. As
depicted, mobile device 116 is in communication with antennas 112
and 114, where antennas 112 and 114 transmit information to mobile
device 116 over a forward link 118 and receive information from
mobile device 116 over a reverse link 120. Moreover, mobile device
122 is in communication with antennas 104 and 106, where antennas
104 and 106 transmit information to mobile device 122 over a
forward link 124 and receive information from mobile device 122
over a reverse link 126. In a frequency division duplex (FDD)
system, forward link 118 can utilize a different frequency band
than that used by reverse link 120, and forward link 124 can employ
a different frequency band than that employed by reverse link 126,
for example. Further, in a time division duplex (TDD) system,
forward link 118 and reverse link 120 can utilize a common
frequency band and forward link 124 and reverse link 126 can
utilize a common frequency band.
[0045] Each group of antennas and/or the area in which they are
designated to communicate can be referred to as a sector of base
station 102. For example, antenna groups can be designed to
communicate to mobile devices in a sector of the areas covered by
base station 102. In communication over forward links 118 and 124,
the transmitting antennas of base station 102 can utilize
beamforming to improve signal-to-noise ratio of forward links 118
and 124 for mobile devices 116 and 122. Also, while base station
102 utilizes beamforming to transmit to mobile devices 116 and 122
scattered randomly through an associated coverage, mobile devices
in neighboring cells can be subject to less interference as
compared to a base station transmitting through a single antenna to
all its mobile devices.
[0046] Base station 102 and mobile devices 116, 122 can be employed
in connection with dynamic clustering in a coordinated multi-point
(CoMP) environment (e.g., network multiple-input multiple-output
(MIMO) environment, . . . ). Dynamic clustering can be utilized to
adapt cooperation strategies to an actual deployment and can be
based upon location and/or priority of active users (e.g., mobile
devices 116, 122, disparate mobile devices (not shown), . . . ),
which can vary over time. Dynamic clustering can mitigate a need
for network planning and cluster boundaries, while potentially
yielding an enhanced throughput/fairness tradeoff.
[0047] In contrast, conventional CoMP approaches typically utilize
cooperation strategies based on predetermined static clustering of
network nodes (e.g., base stations including base station 102, . .
. ). Hence, static master clusters can commonly be chosen based on
assumed network topology such as hexagonal layout or known quality
of backhaul links within master clusters within a Remote Radio Head
context (e.g., Remote Radio Head configurations can include one or
more remote nodes connected to a macro base station via high
quality backhaul links, . . . ). Moreover, interference at
boundaries of master clusters can be handled by traditional
interference management techniques such as, for instance,
fractional reuse, etc. While dynamic cooperative transmissions can
be sent within static clusters, such conventional techniques differ
from approaches set forth herein where clustering strategies are
dynamically selected.
[0048] System 100 can dynamically select clustering strategies in a
CoMP environment. More particularly, base station 102 and disparate
base station(s) can each effectuate distributed decisions to
converge to an optimized set of clusters at a given point in time.
The distributed decisions effectuated by base station 102 and the
disparate base station(s) can be based on a finite order strategy
constraint to limit complexity of inter-site multi-antenna
scheduling and packet sharing. Further, a utility based distributed
negotiation framework based on message passing (e.g., using a
belief propagation framework, . . . ) can be leveraged by base
station 102 and the disparate base station(s) to dynamically yield
the clustering strategy decisions.
[0049] Now turning to FIG. 2, illustrated is an example system 200
that leverages a downlink cooperation framework that employs a
network-wide strategy 202 where a set of base stations in a
deployment cooperatively operate. As depicted, system 200 includes
a set of base stations 204-216 and a set of mobile devices 218-244.
It is contemplated, however, that system 200 can include
substantially any number of base stations and/or substantially any
number of mobile devices and is not limited to the illustrated
example.
[0050] As shown, network-wide strategy 202 can cover all base
stations 204-216 and all mobile devices 218-244 in the deployment.
Thus, base stations 204-216 can cooperate to yield a scheduler
decision where each base station 204-216 can be involved in data
transmission to each mobile device 218-244. For instance, the set
of base stations 204-216 can be scheduled to transmit to a
particular mobile device 218-244, a subset of base stations 204-216
can be scheduled to transmit to a particular mobile device 218-244,
and so forth. Further, scheduler decisions can be based upon a
utility metric. For example, the utility metric can be a function
of weighted rates that can be achieved for different mobile devices
218-244.
[0051] A strategy S can be defined as a set of base stations (e.g.,
nodes, cells, . . . ), mobile devices, underlying antenna weights
and power spectral densities (PSDs) at base stations that serve
mobile devices covered by the strategy S. The set of base stations
covered by strategy S can be referred to as N(S) and the set of
mobile devices covered by strategy S can be referred to as Y(S).
Moreover, a rate achieved by a mobile device y under strategy S at
time t per allocated resource can be R.sub.y,t(S), a utility metric
associated with strategy S at time t can be U.sub.t(S), and a
(relative) priority of mobile device y at time t based on, for
instance, quality of service (QoS), fairness, etc. can be
p.sub.y,t. For example, fairness can be supported by p.sub.y,t
being inversely proportional to an amount of data that mobile
device y has received. According to an example, the utility metric
can be evaluated as follows:
U t ( S ) = .DELTA. y .di-elect cons. Y ( S ) p y , t R y , t ( S )
. ##EQU00001##
[0052] Referring again to system 200, the set of base stations
covered by network-wide strategy 202, N(S 202), includes base
stations 204-216 and the set of mobile devices covered by
network-wide strategy 202, Y(S 202), includes mobile devices
218-244. At a time t, however, scheduling decisions aimed at
maximizing the utility metric U.sub.t(S 202) for network-wide
strategy 202 can be overly complex due to the number of base
stations 204-216 and mobile devices 218-244 covered thereby.
Moreover, it can be impractical and unneeded to involve all base
stations 204-216 in system 200 when serving each mobile device
218-244 in system 200 (e.g., a given mobile device can be impacted
by a finite number of base station(s) from system 200, . . . ).
[0053] Now turning to FIG. 3, illustrated is an example system 300
that employs dynamic clustering based upon a finite order strategy
constraint in a wireless communication environment. Similarly to
the example shown in FIG. 2, system 300 can include base stations
204-216 and mobile devices 218-244; yet, it is to be appreciated
that substantially any number of base stations and/or mobile
devices can be included in system 300. In contrast to the example
of FIG. 2 where network-wide strategy 202 encompassing base
stations 204-216 and mobile devices 218-244 is leveraged, which
results in complex scheduling decisions, system 300 dynamically
forms a plurality of smaller, local strategies 302-310. Thus,
system 300 can include a union of smaller, disjoint strategies
302-310, each with a limited maximum order. Local strategies
302-310 can be manageable in terms of association and spatial
processing complexity. Moreover, intuitively a globally optimal
strategy can include a large number of finite order strategies
(e.g., strategies 302-310, limited order local strategies, . . . )
since high gain long loops on base stations and mobile devices can
be infrequent.
[0054] At any point in time (e.g., for a particular subframe, . . .
), an optimized set of local strategies 302-310 leveraged in system
300 can be dynamically defined (e.g., to yield optimal network-wide
utility, . . . ) to set forth a plurality of groups of cooperating
base station(s) and corresponding mobile device(s) to be served
thereby. Hence, FIG. 3 illustrates the set of local strategies
302-310 dynamically selected for a particular time. As shown for
the particular time, local strategy 302 can cover base stations 204
and 206 and mobile devices 220 and 222, local strategy 304 can
cover base station 208 and mobile device 234, local strategy 306
can cover base stations 210 and 214 and mobile devices 226 and 228,
local strategy 308 can cover base station 212 and mobile device
232, and local strategy 310 can cover base station 216 and mobile
device 244. At a different time, a differing optimized set of local
strategies, each covering a corresponding subset of base stations
204-216 and corresponding subset of mobile devices 218-244, can be
chosen.
[0055] Local strategies 302-310 can each correspond to a cluster
including a limited number of base station(s) (e.g., from the set
of base stations 204-216, . . . ) and mobile device(s) (e.g., from
the set of mobile devices 218-244, . . . ). Further, each of the
clusters can effectuate its own scheduling. Base stations included
in a common cluster can be scheduled to effectuate various
cooperation techniques as described herein.
[0056] A strategy order can be defined as a number of base stations
involved in a given strategy (e.g., local strategy, . . . ). For
instance, strategy order can be referred to as |N(S)| (e.g.,
cardinality of the set N(S), number of members of the set N(S), . .
. ), and |N(S)| can be a member of a set {1, . . . , X.sub.S}
(e.g., |N(S)|.epsilon.{1, . . . , X.sub.S}, . . . ). Further,
X.sub.S is a maximum order that can be allowed in system 300.
According to an example, X.sub.S can be 3. By way of another
example, X.sub.S can be 2. Yet, it is to be appreciated that
X.sub.S can be any integer greater than 3 and is not limited to the
aforementioned examples. As shown, local strategies 304, 308, and
310 can each include one respective base station, and thus, can be
first order strategies. Moreover, local strategies 302 and 306 can
each include two respective base stations, and hence, can be second
order strategies. It is to be appreciated, however, that system 300
can also support third order strategies (or higher order
strategies) dependent upon a value of X.sub.S.
[0057] A first order strategy (e.g., local strategy 304, local
strategy 308, local strategy 310, . . . ) can be similar to a
classic wireless communication model that lacks coordination
between base stations. Hence, a mobile device included in a first
order strategy can be served by a base station included in the
first order strategy. In contrast, a second order strategy (e.g.,
local strategy 302, local strategy 306, . . . ) can leverage
cooperation between base stations included in such strategy. Thus,
mobile devices covered by a second order strategy can be served by
two base stations included in the second order strategy in a
cooperative manner.
[0058] According to an example, second order strategy 302 can
include two base stations 204 and 206, each of which can
respectively have one transmit antenna. Mobile devices 220 and 222
covered by second order strategy 302 can be cooperatively served by
the two base stations 204 and 206. Hence, virtual MIMO can be
carried out within second order strategy 302, effectively treating
the two base stations 204 and 206 as one base station with two
antennas, by leveraging the two transmit antennas associated with
the two base stations 204 and 206. However, it is to be appreciated
that the claimed subject matter is not limited to the foregoing
example.
[0059] Second order strategies and higher order strategies can
enable base stations to pool together resources, antennas, and the
like. Further, such strategies can allow for joint scheduling
handled by base stations included in a common local strategy.
Moreover, information can be shared between base stations in the
common local strategy. For instance, the shared information can
include channel information (e.g., for channel(s) between base
station(s) and mobile device(s) in the local strategy, . . . ),
packets (e.g., to be transmitted from one or more base stations in
the local strategy, . . . ), and so forth. Hence, within each local
strategy 302-310, base station(s) and/or mobile device(s) can
cooperate with each other (e.g., to yield coordinated scheduling
decisions, . . . ); yet, base station(s) and/or mobile device(s)
need not cooperate with base station(s) and/or mobile device(s)
included in differing local strategies 302-310 (e.g., cooperation
need not extend across local strategies 302-310, . . . ). Further,
each local strategy 302-310 can assess interference caused by other
local strategies 302-310 and/or attempt to mitigate an impact of
such interference.
[0060] An overall strategy S within system 300 can be a direct sum
of local strategies, where strategy order of the local strategies
can be constrained to a maximum value (e.g.,
S = .sym. l S l with N ( S l ) .ltoreq. X S ##EQU00002##
where l is a local strategy index, . . . ). For example, each base
station can be included in at most only one local strategy at a
given time; thus, an intersection of a set of base stations covered
by a first local strategy (e.g., with an index l, . . . ) and a set
of base stations covered by a second local strategy (e.g., with an
index l' for all differing l and l', . . . ) at a given time is an
empty set (.phi.) (e.g.,
N(S.sub.l).andgate.N(S.sub.l')=.phi..A-inverted.l.noteq.l', . . .
). Moreover, an overall utility at a given time t can be a sum of
utilities corresponding to the local strategies at the given time t
(e.g.,
U t ( S ) = l U t ( S l ) , ) . ##EQU00003##
Each of the utilities corresponding to the local strategies can be
evaluated as:
U t ( S l ) = .DELTA. y .di-elect cons. Y ( S l ) p y , t R y , t (
S l ) . ##EQU00004##
It is to be appreciated, however, that the claimed subject matter
is not limited to the foregoing example, and rather, it is
contemplated that, pursuant to another example, a base station can
concurrently be included in more than one local strategy.
[0061] Subsets of base stations 204-216 and subsets of mobile
devices 218-244 are dynamically grouped over time to yield the time
varying set of local strategies 302-310. In contrast, conventional
techniques that allow grouping of base stations typically define
static clusters, which remain constant over time (e.g., the same
base stations are grouped together over time, . . . ). Since system
300 leverages dynamic clustering, various conditions such as
positioning of mobile devices 218-244, buffer levels of mobile
devices 218-244, channel conditions between base station(s) 204-216
and mobile device(s) 218-244, and the like can be considered when
forming local strategies 302-310 at a given time. Further, at a
next time, a differing set of local strategies can be formed (e.g.,
depending on changes to the various conditions within system 300, .
. . ). Thus, for example, while local strategy 302 includes base
stations 204 and 206 and mobile devices 220 and 222 at a particular
time in the depicted example of FIG. 3, at a next time a local
strategy can be selected that includes base stations 204 and 210
and mobile devices 220 and 222, while base station 206 can be
covered by a differing local strategy. Moreover, following this
example, at a further subsequent time, a local strategy can be
chosen that groups base station 204 and mobile device 218, while
base stations 206 and 210 and mobile devices 220 and 222 can be
included in one or more differing local strategies. However, it is
to be appreciated that the claimed subject matter is not limited to
the foregoing example.
[0062] Referring to FIG. 4, illustrated is a system 400 that
employs distributed strategy negotiation in a wireless
communication environment. System 400 includes a base station 402
and a plurality of disparate base stations 404. Further, although
not shown, it is contemplated that system 400 can include
substantially any number of mobile devices. Base station 402 can
interact with at least a subset of disparate base stations 404 to
transmit and/or receive information, signals, data, instructions,
commands, bits, symbols, and the like. Moreover, based upon the
interaction, base station 402 and disparate base stations 404 can
each select a respective local strategy to implement from a
respective set of possible local strategies. Base station 402 and
disparate base stations 404 can converge to a compatible set of
local strategies across system 400 that yield clusters which are
non-overlapping.
[0063] Base station 402 can further include a clustering component
406, a metric evaluation component 408, and a negotiation component
410. Similarly, although not shown, it is contemplated that
disparate base stations 404 can each likewise include a respective
clustering component (e.g., similar to clustering component 406, .
. . ), a respective metric evaluation component (e.g., similar to
metric evaluation component 408, . . . ), and a respective
negotiation component (e.g., similar to negotiation component 410,
. . . ).
[0064] Clustering component 406 can dynamically select a local
strategy to implement with base station 402 from the set of
possible local strategies. For instance, clustering component 406
can choose to form a cluster with a particular one (or subset) of
disparate base stations 404 at a given time based upon the selected
local strategy. Moreover, one or more mobile devices can be
included in the cluster corresponding to the selected local
strategy at the given time. Further, at a next time, clustering
component 406 can, but need not, elect to utilize a differing local
strategy from the set of possible local strategies. Moreover, each
of disparate base stations 404 can similarly dynamically select a
respective local strategy to leverage as a function of time. Thus,
system 400 supports effectuating a fully distributed strategy
determination, where each base station (e.g., base station 402,
each of disparate base stations 404, . . . ) can evaluate possible
local strategies in which the base station can be involved to
select a particular local strategy for the base station at a given
time.
[0065] For each base station to select the particular local
strategy to implement at a given time, each base station can
evaluate a metric. More particularly, metric evaluation component
408 (and similar metric evaluation components of disparate base
stations 404) can evaluate marginal utilities (e.g., local
utilities, . . . ) of possible local strategies in which base
station 402 can cooperate with neighbor base stations (e.g., one or
more of disparate base stations 404, . . . ). A marginal utility
(e.g., local utility, . . . ) analyzed by metric evaluation
component 408 can be a utility of a local strategy in isolation
from a remainder of a network. Moreover, neighbor base stations and
base station 402 can share channel state information (CSI) and/or
information concerning priority of common mobile devices; the
shared information can be used by metric evaluation component 408
(and similar metric evaluation components of disparate base
stations 404) to effectuate analyzing the marginal utilities (e.g.,
local utilities, . . . ).
[0066] Further, upon metric evaluation component 408 yielding the
marginal utilities (e.g., local utilities, . . . ) associated with
the possible local strategies, negotiation component 410 can employ
message passing to agree on a compatible set of local strategies
(e.g., marginal strategies, . . . ) across system 400. For
instance, message passing can be effectuated across base stations
in system 400 (e.g., base station 402 and disparate base stations
404, . . . ). Moreover, base stations in system 400 can exchange
strategy and utility information with respective neighbors through
message passing. By way of example, base station 402 can exchange
strategy and utility information with its neighbor base stations
(e.g., subset of disparate base stations 404, . . . ). Moreover, it
is contemplated that the message passing can be iterative; however,
the claimed subject matter is not so limited. Message passing can
be implemented in system 400 to enable each base station to compute
an estimate of an overall network-wide utility associated with a
particular marginal strategy (e.g., particular local strategy from
the set of possible local strategies associated with the base
station, . . . ). Moreover, message passing effectuated in system
400 can be analogous to message passing decoding, wherein
iterations lead to a symbol-wise metric that can reflect value and
reliability of a bit within a globally optimal solution. Further,
utility based quantities for inter-base station exchange can be
generalized to account for additional (practical) constraints such
as backhaul quality, preferred cooperation technique(s), and the
like.
[0067] Negotiation component 410 can enable base station 402 to
transmit utility information to and received utility information
from neighbor base stations (e.g., subset of disparate base
stations 404, . . . ). By exchanging utility information, base
station 402 and disparate base stations 404 can converge to a set
of clusters to be employed in system 400 (e.g., by base station 402
and disparate base stations 404 each selecting respective local
strategies that maximize overall network-wide utility, . . . ),
where the clusters in the set are non-contradictory (e.g., at any
point in time on any resource the clusters are non-overlapping, . .
. ). Pursuant to an example, if clustering component 406 of base
station 402 selects a local strategy (e.g., second order local
strategy, . . . ) at a given time where base station 402 and a
specific one of disparate base stations 404 are clustered, then the
specific one of disparate base stations 404 (e.g., disparate
clustering component thereof, . . . ) elects a local strategy
(e.g., second order local strategy, . . . ) at the given time where
the specific one of disparate base stations 404 and base station
402 are clustered while remaining disparate base stations 404 do
not select respective local strategies that involve base station
402 or the specific one of disparate base stations 404 at the given
time.
[0068] Base station 402 and disparate base stations 404 can enable
controlling the set of local strategies chosen to be utilized in
system 400 in a distributed fashion rather than employing a
centralized controller. Base station 402 and disparate base
stations 404 can each consider respective sets of possible local
strategies (e.g., base station 402 can analyze utilities associated
with each of the possible local strategies using metric evaluation
component 408, disparate base stations 404 can similarly evaluate
utilities, . . . ). Moreover, message passing can be effectuated
(e.g., with negotiation component 410 and similar negotiation
components of disparate base stations 404, . . . ) to exchange
utility information between neighbors, which can lead to base
station 402 and disparate base stations 404 forming a convergent
solution across system 400 at a given time.
[0069] Due to the exchange of utility information effectuated by
negotiation component 410, metric evaluation component 408 can
yield an estimate of network-wide utility for the possible local
strategies in which base station 402 can be involved. Thus, metric
evaluation component 408 can compute weighted sum rates of mobile
devices that are involved in each of the possible local strategies
as well as estimate overall sum rates across an entire network
conditioned on the fact that each of the possible local strategies
are employed by base station 402. Hence, base station 402 and
disparate base stations 404 can each estimate network-wide utility
conditioned upon each possible local strategy the base stations can
respectively leverage.
[0070] Turning to FIG. 5, illustrated is an example system 500 that
employs message passing in a wireless communication environment.
System 500 includes a node 0 502 and three nodes 504, 506, and 508
(e.g., node 1 504, node 2 506, and node 3 508, . . . ) that
neighbor node 0 502. Nodes 502-508 can also be referred to as base
stations 502-508. Further, each node 502-508 can be substantially
similar to base station 402 of FIG. 4.
[0071] By way of example, node 0 502 can cooperate with one of node
1 504, node 2 506, or node 3 508 (e.g., a constraint can be
utilized within system 500 to limit a number of nodes 502-508 and
mobile devices that can be included within a common cluster,
assuming that a maximum order strategy supported in system 500 is
two, . . . ). For instance, cooperation between node 0 502 and node
1 504 can result in a certain local utility, which is a weighted
sum rate across mobile devices served by such local strategy.
Further, if node 0 502 cooperates with node 1 504, then node 0 502
can be unable to cooperate with node 2 506 or node 3 508. Based
upon a measure of local utility (e.g., yielded by metric evaluation
component 408 of FIG. 4, . . . ), node 0 502 can recognize that
cooperation with node 1 504 yields a higher local utility in
comparison to cooperation with node 2 506 or node 3 508; however,
cooperation between node 0 502 and node 1 504 can be detrimental to
overall network-wide utility as compared to node 0 502 operating
under a differing local strategy. Thus, message passing can be
employed to propagate messages across system 500, where such
messages allow nodes 502-508 to each estimate network-wide utility
conditioned upon each of the possible local strategies that each of
nodes 502-508 can respectively implement. For instance, after a
number of iterations, node 0 502 can estimate network-wide utility
associated with each possible local strategy that can be selected
by node 0 502, and node 0 502 can choose a particular local
strategy with a maximum estimate of network-wide utility. Hence,
the aforementioned message passing algorithm can enable converging
to a global optimal solution. Moreover, nodes 502-508 can
dynamically decide upon local strategies over time subject to
channel conditions, mobile device conditions, and the like. It is
to be appreciated, however, that the claimed subject matter is not
limited to the foregoing example.
[0072] FIG. 5 shows a cooperation graph with vertices represented
by nodes and edges represented by (potential) cooperation
relationships. For instance, an edge can be between two nodes
(e.g., node a and node b, . . . ) at a given time if there exists a
common mobile device with active priorities, wherein the common
mobile device receives pilots from both nodes (e.g., strengths of
received pilots can be similar to an active or candidate set
concept, . . . ). Thus, two nodes that have a common edge can be
referred to as neighbors.
[0073] Utility information can be passed between neighbors in
system 500 as part of the distributed negotiation framework
described herein. For instance, node 0 502 can pass utility
information to each of its neighbors (e.g., nodes 504-508, . . . )
and can receive utility information from each of its neighbors
(e.g., by employing negotiation component 410 of FIG. 4, . . . ).
In the examples described below, the utility information can be
transmitted from node p (e.g., source node, . . . ) to node q
(e.g., target node, . . . ); for instance, utility information can
be sent from node 0 502 to node 1 504, from node 1 504 to node 0
502, and so forth.
[0074] According to various embodiments, extrinsic utilities
transmitted from node p to node q as part of a distributed
negotiation framework can include a cooperative utility value and a
non-cooperative utility value. The cooperative utility value can be
referred to as U.sub.p,q.sup.(c) and the non-cooperative utility
value can be referred to as U.sub.p,q.sup.(n). The cooperative
utility value transmitted from node p to node q can reflect an
estimate of a total utility of a sub-graph connected to node q
through node p assuming that node p is not involved in any strategy
that excludes node q, hence allowing potential cooperation with
node q. Further, the non-cooperative utility value transmitted from
node p to node q can reflect an estimate of a total utility of a
sub-graph connected to node q through node p assuming that node p
is involved in a strategy that excludes node q; thus, node p
potentially does not cooperate with node q. Moreover, an implicit
assumption can be that sub-graphs are non-overlapping. Additionally
or alternatively, this message passing algorithm can assume a lack
of loops; however, the claimed subject matter is not so
limited.
[0075] As part of the foregoing distributed negotiation framework,
L.sub.p represents a set of indexes of all potential marginal
strategies (e.g., potential local strategies, . . . ) associated
with node p. Moreover, U.sub.p,t(S.sub.l) can be an estimate of
network-wide sum utility (NWSU) conditioned on the marginal
strategy S.sub.l that involves node p computed by node p at time t.
Further, l can be a member of L.sub.p (e.g., l.epsilon.L.sub.p,
L.sub.p={1, 2, 3} in the example shown in FIG. 5, . . . ).
Accordingly, the estimate of the network-wide sum utility
conditioned on a particular marginal strategy can be the sum of a
utility for the marginal strategy plus a sum-utility of all
sub-graphs connected via cooperative nodes plus a sum-utility of
all sub-graphs connected via non-cooperative nodes, which can be
represented as follows:
U p , t ( S l ) = U t ( S l ) + m .di-elect cons. N ( S l ) U m , p
( c ) + m U m , p ( n ) . ##EQU00005##
[0076] Accordingly, node p can identify an index (l.sub.p,q,t) of a
best strategy involving node p and q at time t as follows:
l p , q , t = arg max l .di-elect cons. L p , q .di-elect cons. N (
S l ) U p , t ( S 1 ) ##EQU00006##
U p , q ( c ) := U p , t ( S l p , q , t ) - U t ( S l p , q , t )
- U q , p ( c ) ##EQU00007## U p , q ( c ) := m .di-elect cons. N (
S l p , q , t ) m .noteq. q U m , p ( c ) + m N ( S l p , q , t ) U
m , p ( n ) ##EQU00007.2##
The foregoing can represent a sum utility of all sub-graphs
connected to node p except sub-graphs connected through node q
assuming cooperation between node p and node q. Moreover, node p
can recognize an index (l'.sub.p,q,t) of a best strategy involving
node p but not node q at time t as follows:
l p , q , t ' = arg max l .di-elect cons. L p , q N ( S l ) U p , t
( S 1 ) ##EQU00008## U p , q ( n ) := U p , t ( S l p , q , t ' ) -
U q , p ( n ) ##EQU00008.2## U p , q ( c ) := m .di-elect cons. N (
S l p , q , t ) m .noteq. q U m , p ( c ) + m N ( S l p , q , t ) U
m , p ( n ) ##EQU00008.3##
The above can correspond to a sum utility of all sub-graphs
connected to node p except sub-graphs connected through node q
assuming no cooperation between node p and node q. The utility of
marginal strategy S.sub.l can be evaluated as follows:
U t ( S l ) = y .di-elect cons. Y ( S l ) p y , t R y , t ( S l ) .
##EQU00009##
Further, marginal strategy selection can be effectuated according
to
l p , t * = arg max l .di-elect cons. L p U p , t ( S l ) ,
##EQU00010##
which can be analogous to a hard decision at an end of message
passing decoding.
[0077] Below is an example extrinsic utility calculation that can
be performed by node 0 502. Node 0 502 can calculate
U.sub.0,1.sup.(c) and U.sub.0,1.sup.(n) given values of
U.sub.1,0.sup.(c), U.sub.1,0.sup.(n), U.sub.2,0.sup.(c),
U.sub.2,0.sup.(n), U.sub.3,0.sup.(c), and U.sub.3,0.sup.(n). When
evaluating U.sub.0,1.sup.(c), S.sub.1 is the possible cooperative
strategy between node 0 502 and node 1 504, and thus,
l.sub.p,q,t=1. It thus follows that
U.sub.0,1.sup.(c)=U.sub.0,t(S.sub.1)-U.sub.t(S.sub.1)-U.sub.1,0.sup.(c).
Moreover, the estimate of network-wide sum utility conditioned on
the marginal strategy S.sub.1 that involves node 1 504 computed by
node 0 502 can be identified as
U.sub.0,t(S.sub.1)=U.sub.t(S.sub.1)+U.sub.1,0.sup.(c)+U.sub.2,0.sup.(n)+U-
.sub.3,0.sup.(n). Hence, U U.sub.0,1.sup.(c) can equal a sum
utility over a part of an overall network connected to node 1 504
through node 0 502 assuming cooperation between node 0 502 and node
1 504, where node 2 506 and node 3 508 do not cooperate with node 0
502 (e.g., U.sub.0,1.sup.(c)=U.sub.2,0.sup.(n)+U.sub.3,0.sup.(n), .
. . ). Further, when node 0 502 analyzes U.sub.0,1.sup.(n),
strategies S.sub.2 and S.sub.3 can be options for node 0 502
assuming no cooperation with node 1 504. For instance, it can be
assumed that S.sub.2 is a better strategy for node 0 502 (e.g.,
U.sub.0,t(S.sub.2)>U.sub.0,t(S.sub.3), . . . ), and then
l'.sub.p,q,t=2 and
U.sub.0,1.sup.(n)=U.sub.0,t(S.sub.2)-U.sub.1,0.sup.(n). Moreover,
the following can be yielded:
U.sub.0,t(S.sub.2)=U.sub.t(S.sub.2)+U.sub.2,0.sup.(c)+U.sub.1,0.sup.(n)+U-
.sub.3,0.sup.(n)Accordingly, U.sub.0,1.sup.(n) can equal a sum
utility over a part of an overall network connected to node 1 504
through node 0 502 assuming no cooperation between node 0 502 and
node 1 504 (e.g., hence node 0 502 considers the best cooperation
other than with node 1 504 which by assumption is cooperation with
node 2 506 via strategy S.sub.2, . . . ); thus,
U.sub.0,1.sup.(n)=U.sub.t(S.sub.2)+U.sub.2,0.sup.(c)+U.sub.3,0.sup.(n).
It is to be appreciated, however, that the claimed subject matter
is not limited to the foregoing example.
[0078] According to other embodiments, extrinsic utilities
transmitted from node p to node q as part of a distributed
negotiation framework can include a plurality of utility values.
For instance, if node p considers T possible local strategies, then
node p can send T messages, where T can be substantially any
integer; yet, the claimed subject matter is not so limited. A
projected utility of a strategy S.sub.l can be computed as a sum of
a local utility (e.g., U.sub.t(S.sub.l), . . . ) and extrinsic
utilities from neighbor nodes that are compatible with S.sub.l. The
extrinsic utility value transmitted from node p to node q can be
referred to as U.sub.p,q.sup.(m), which reflects the total utility
of a sub-graph connected to node q through node p under constraints
on node q reported to node q. Moreover, different messages m can
represent different constraints on node q. For example, node p can
be involved in cooperation with node q; following this example,
such extrinsic utility can be added up by node q to compute
projected utility of a strategy where node q cooperates with node
p. By way of another example, node p can lack cooperation with node
q; accordingly, the extrinsic utility can be added up by node q to
compute a projected utility of a strategy where node q does not
cooperate with node p. Further, the messages m can indicate nodes
and/or mobile devices involved in cooperation under a particular
strategy so that node q does not add up extrinsic utilities
referring to any node and/or mobile device to a local utility of
its own strategy that involves the same node and/or mobile
device.
[0079] As part of the above noted distributed negotiation
framework, L.sub.p represents a set of indexes of all potential
local strategies (e.g., potential marginal strategies, . . . )
associated with node p. Moreover, U.sub.p,t(S.sub.l) can be a
projected network-wide sum utility (NWSU) conditioned on the local
strategy S.sub.l that involves node p computed by node p at time t.
Further, l can be a member of L.sub.p (e.g., l.epsilon.L.sub.p,
L.sub.p={1, 2, 3} in the example shown in FIG. 5, . . . ).
Accordingly, a projection of a network-wide sum utility conditioned
on a particular local strategy S.sub.l can be calculated as
follows:
U p , t ( S l ) = U t ( S l ) + q max 1 .ltoreq. m .ltoreq. M q
.xi. ( m ) ( S l , q , p ) U q , p ( m ) M q ##EQU00011##
can represent a total number of messages passed from node q to node
p, U.sub.p,q.sup.(m) can represent the m-th extrinsic utility
received from node q, and .xi..sup.(m)(S.sub.l,q,p) can be a
compatibility verification for the m-th message from node q to node
p with strategy S.sub.l which can have a value of 0 or 1. Further,
local strategy selection can be effectuated according to
l p , t * = arg max l .di-elect cons. L p U p , t ( S l ) ,
##EQU00012##
which can be analogous to a hard decision at an end of message
passing decoding.
[0080] The total number of messages M.sub.q can match a number of
constraints corresponding to possible local strategies of a target
node. For instance, for each common neighbor of a source node and a
target node, a message can be added that corresponds to no
cooperation between the source node and that common neighbor (e.g.,
a number of common neighbors equals a number of messages, . . . ).
By way of another illustration, messages that correspond to
cooperation with one common neighbor and non-cooperation with
another common neighbor can be considered if second order or higher
strategies are leveraged (e.g., a number of messages equals the
number of common neighbors multiplied by the number of common
neighbors minus one where the product can be divided by two, . . .
). Further, a message that does not involve any common neighbors
can yield no constraints on a target node, and thus, can be used
with any non-cooperation strategy. Moreover, it can be unnecessary
to use multiple messages under a common set of constraints; rather,
the source (or target) node can select a message with a highest
utility under the given constraints. Pursuant to an example,
extrinsic messages chosen for sending can include one message that
does not involve the target or any common neighbors (e.g.,
identified via knowledge of a common neighbor list, . . . ) and
remaining messages selected with maximum extrinsic utilities; yet,
it is to be appreciated that the claimed subject matter is not so
limited.
[0081] Below are example projected utility calculations that
leverage the above noted plurality of extrinsic utilities. It is to
be appreciated, however, that the claimed subject matter is not so
limited.
[0082] According to an illustration, each node p (e.g., node 1 504,
node 2 506, node 3 508, . . . ) can have one possible constraint
for node 0 502, namely that node 0 502 is involved in cooperation
with node p, where 1.ltoreq.p.ltoreq.3 for system 500. Hence, node
p can send the following two messages to node 0 502:
U.sub.p,0.sup.(1) which is conditioned on node p cooperating with
node 0 502 and U.sub.p,0.sup.(2) which is conditioned on node p not
cooperating with node 0 502. For instance, node 0 502 can support
three possible local strategies (e.g., S.sub.1 where node 0 502 can
be clustered with node 1 504, S.sub.2 where node 0 502 can be
clustered with node 2 506, or S.sub.3 where node 0 502 can be
clustered with node 3 508, . . . ). Thus, node 0 502 can evaluate
projected network-wide sum utilities for each of the three possible
local strategies per the below:
U.sub.0,t(S.sub.1)=U.sub.t(S.sub.1)+U.sub.1,0.sup.(1)+U.sub.2,0.sup.(2)+-
U.sub.3,0.sup.(2)
U.sub.0,t(S.sub.2)=U.sub.t(S.sub.2)+U.sub.1,0.sup.(2)+U.sub.2,0.sup.(1)+-
U.sub.3,0.sup.(2)
U.sub.0,t(S.sub.3)=U.sub.t(S.sub.3)+U.sub.1,0.sup.(2)+U.sub.2,0.sup.(2)+-
U.sub.3,0.sup.(1)
[0083] Moreover, node 0 502 can compute extrinsic utilities, which
can be transmitted to a target neighboring node (e.g., node 1 504,
node 2 506, node 3 508, . . . ). The extrinsic utility passed to
the target neighboring node can represent a fraction of a
network-wide utility that excludes utility of the target
neighboring node and other nodes connected to a source node (e.g.,
node 0 502, . . . ) through the target. This can imply a loop-less
network graph, where a node has at most one path to another node.
Although loop-less graphs typically don't exist, belief propagation
algorithms can be designed under such assumption (e.g., short loops
can have more impact than long loops, there can be fewer short(er)
loops compared to long(er) loops, . . . ).
[0084] For a set .OMEGA..sub.p,q.sup.(m) of nodes that can
cooperate with node q decided by node p for a message m to be sent
to node q, a strategy that maximizes projected utility under the
stated constraints can be evaluated as follows:
l * = arg max l .di-elect cons. L p : N ( S l ) .OMEGA. p , q ( m )
= .phi. U p , t ( S l ) . ##EQU00013##
Moreover, the extrinsic utility for the target node q can be
obtained as this projected utility less the extrinsic contribution
to this utility from the target node q, as shown below:
U p , q ( m ) = U t ( S l * ) + q ' .noteq. q max 1 .ltoreq. m
.ltoreq. M q ' .xi. ( m ) ( S l * , q ' , p ) U q ' , p ( m )
##EQU00014##
In general, the source node p can compute and send to the target
node q multiple extrinsic utilities U.sub.p,q.sup.(m) corresponding
to different sets .OMEGA..sub.p,q.sup.(m). Although at least one
message can be used for every possible set .OMEGA..sub.p,q.sup.(m),
the total number of messages passed from node p to node q can be
pruned without much loss in performance. For example, pruning can
be accomplished by selecting a limited number of messages with the
largest values of corresponding projected utilities.
[0085] According to the above illustration with one constraint,
node 1 504 can have one possible constraint for node 0 502, namely
that node 1 504 is involved in cooperation with node 0 502. Thus,
node 0 502 sends two messages to node 1 504: U.sub.0,1.sup.(1)
which corresponds to node 1 504 cooperating with node 0 502 and
U.sub.0,1.sup.(2) which corresponds to node 1 504 not cooperating
with node 0 502. Node 0 502 can evaluate the extrinsic utilities as
follows:
U.sub.0,1.sup.(1)=0
If U.sub.0,t(S.sub.2)>U.sub.0,t(S.sub.3), then
U.sub.0,1.sup.(2)=U.sub.t(S.sub.2)+U.sub.2,0.sup.(1)+U.sub.3,0.sup.(2)=U-
.sub.0,t(S.sub.2)-U.sub.1,0.sup.(2)
Else,
U.sub.0,1.sup.(2)=U.sub.t(S.sub.3)+U.sub.2,0.sup.(2)+U.sub.3,0.sup.(1)=U-
.sub.0,t(S.sub.3)-U.sub.1,0.sup.(2)
[0086] FIG. 6 depicts another system 600 that employs message
passing in a wireless communication environment. As shown, node 0
502 can evaluate three possible local strategies: S.sub.1 602 where
node 0 502 and node 1 504 are clustered (e.g., cooperate, . . . ),
S.sub.2 604 where node 0 502 and node 2 506 are clustered (e.g.,
cooperate, . . . ), and S.sub.3 606 where node 0 502 does not
cooperate with either node 1 504 or node 2 506. Moreover, node 1
504 and node 2 506 can be neighbors of each other. By way of
another illustration, each node p (e.g., node 1 504, node 2 506, .
. . ) can have two constraints for node 0 502. The two constraints
can be cooperation with node 0 502 and cooperation with another
neighbor of node 0 502. Hence, node p can sent three messages to
node 0 502: U.sub.p,0.sup.(1) which is conditioned on node p
cooperating with node 0 502, U.sub.p,0.sup.(2) which is conditioned
on node p cooperating with a neighbor node q of node 0 502 (e.g.,
q=2 for p=1, q=1 for p=2, . . . ), and U.sub.p,0.sup.(3) which is
conditioned on node p not cooperating with node 0 502 or a neighbor
of node 0 502. Thus, node 0 502 can evaluate projected network-wide
sum utilities for each of the three possible local strategies
602-606 as follows:
U.sub.0,t(S.sub.1)=U.sub.t(S.sub.1)+U.sub.1,0.sup.(1)+U.sub.2,0.sup.(3)
U.sub.0,t(S.sub.2)=U.sub.t(S.sub.2)+U.sub.1,0.sup.(3)+U.sub.2,0.sup.(1)
U.sub.0,t(S.sub.3)=U.sub.t(S.sub.3)+max{U.sub.1,0.sup.(2),U.sub.1,0.sup.-
(3)}+max{U.sub.2,0.sup.(2),U.sub.2,0.sup.(3)}
[0087] Further, node 0 502 can compute extrinsic utilities, which
can be sent to a target neighboring node (e.g., node 1 504, node 2
506, . . . ). According to the above illustration with two
constraints, node 1 504 can have two possible constraints for node
0 502, namely that node 1 504 is involved in cooperation with node
0 502 and that node 1 504 is involved in cooperation with node 2
506 (e.g., which is also a neighbor of node 0 502, . . . ). Thus,
node 0 502 sends three messages to node 1 504: U.sub.0.sup.(1)
which corresponds to node 1 504 cooperating with node 0 502,
U.sub.0,1.sup.(2) which corresponds to node 1 504 cooperating with
node 2 506, and U.sub.0,1.sup.(3) which corresponds to node 1 504
cooperating with neither node 0 502 nor node 2 506. Node 0 502 can
evaluate the extrinsic utilities as follows:
U.sub.0,1.sup.(1)=0
U.sub.0,1.sup.(2)=U.sub.t(S.sub.3)+max{(U.sub.2,0.sup.(2),U.sub.2,0.sup.-
(3)}=U.sub.0,t(S.sub.3)-max{U.sub.1,0.sup.(2),U.sub.1,0.sup.(3)}
If U.sub.0,t(S.sub.2)>U.sub.0,t(S.sub.3), then
U.sub.0,1.sup.(3)=U.sub.t(S.sub.2)+U.sub.2,0.sup.(1)=U.sub.0,t(S.sub.2)--
U.sub.1,0.sup.(3)
Else,
U.sub.0,1.sup.(3)=U.sub.0,1.sup.(2)
[0088] Now turning to FIG. 7, illustrated is a system 700 that
supports cooperation within clusters in a wireless communication
environment. System 700 includes base station 402, cooperating base
station(s) 702, and non-cooperating base station(s) 704 (e.g.,
cooperating base station(s) 702 and non-cooperating base station(s)
704 can each be substantially similar to base station 402, . . . ).
For instance, cooperating base station(s) 702 and non-cooperating
base station(s) 704 can be disparate base stations 404 of FIG. 4.
As described herein, at a given time, base station 402 and
cooperating base station(s) 702 can dynamically form a cluster 706.
Thus, base station 402 and cooperating base station(s) 702 can
cooperate with each other at the given time; meanwhile, at the
given time, base station 402 and cooperating base station(s) 702 do
not cooperate with non-cooperating base station(s) 704. Moreover,
non-intersecting subsets of non-cooperating base station(s) 704 can
similarly form respective, non-overlapping clusters in which
cooperation can be effectuated. Further, cluster 706 can include
mobile device(s) 708, which are served by base station 402 and
cooperating base station(s) 702. Likewise, although not shown,
system 700 can include mobile devices not included in cluster 706
that are each covered by a respective one of the non-overlapping
clusters dynamically formed by the non-cooperating base station(s)
704 at the given time.
[0089] As described herein, base station 402 can leverage
clustering component 406, metric evaluation component 408, and
negotiation component 410 to dynamically select to cooperate with
cooperating base station(s) 702 at the given time in a distributed
manner. Moreover, base station 402 can include a cooperation
component 710 that can coordinate operation of base station 402 and
cooperating base station(s) 702 to effectuate one or more
cooperation techniques. Hence, upon forming cluster 706,
cooperation component 710 (and similar cooperation component(s) of
cooperating base station(s) 702) can control operations within
cluster 706 to take advantage of cooperation there between.
[0090] With reference to FIGS. 8-10, illustrated are various
example cooperation techniques that can be implemented within a
cluster in a wireless communication environment. For instance, each
of the example cooperation techniques can be managed, scheduled,
coordinated, etc. by respective cooperation components (e.g.,
cooperation component 710 of FIG. 7, . . . ) of base stations
included in each cluster. Depicted are examples of inter-site
packet sharing, cooperative beamforming, and cooperative silence;
it is to be appreciated, however, that the claimed subject matter
is not limited to the examples shown in FIGS. 8-10 as these
techniques are shown for illustration purposes.
[0091] Turning to FIG. 8, illustrated is an example system 800 that
employs inter-site packet sharing (ISPS) (e.g., coherent ISPS, . .
. ) within a cluster 802 in a wireless communication environment.
Cluster 802 includes base stations 804 and 806 and mobile devices
808 and 810 (e.g., second order strategy, . . . ). Inter-site
packet sharing can also be referred to as joint processing or joint
transmission. When leveraging inter-site packet sharing, each base
station 804-806 within cluster 802 can be involved in data
transmission to each mobile device 808-810 included in cluster
802.
[0092] Inter-site packet sharing can be most efficient with a
limited number of transmit antennas per base station 804-806 (e.g.,
limited number of transmit antennas per node, . . . ). For example,
base stations 804-806 can each include one transmit antenna. Thus,
the two base stations 804-806 within cluster 802 can effectively be
utilized as one base station with two antennas when serving mobile
devices 808-810; however, the claimed subject matter is not so
limited.
[0093] Inter-site packet sharing can leverage a high bandwidth
backhaul between base stations 804-806. Moreover, fast
Acknowledgement and Negative Acknowledgement ((N)ACK) distribution
across cooperating base stations 804-806 can be used in system 800.
Further, inter-site packet sharing can be sensitive to channel
state information (CSI). Inter-site packet sharing can be used by a
collection of base stations 804-806 and mobile devices 808-810 that
yield a substantial performance benefit.
[0094] Now referring to FIG. 9, illustrated is an example system
900 that implements cooperative beamforming within a cluster 902 in
a wireless communication environment. Cluster 902 includes base
stations 904 and 906 and mobile devices 908 and 910 (e.g., second
order strategy, . . . ). Cooperative beamforming can also be
referred to as coordinated beamforming or distributed beamforming
(DBF). To effectuate cooperative beamforming, base stations 904-906
can each have multiple transmit antennas; yet, the claimed subject
matter is not so limited.
[0095] As depicted, base station 904 can serve mobile device 910
and base station 906 can serve mobile device 908 within cluster
902. When base station 904 sends a transmission to mobile device
910, base station 904 can yield a beam that mitigates interference
to mobile device 908 (e.g., beams to mobile device 910 with
transmit nulling to mobile device 908, . . . ). Thus, each base
station 904-906 can coordinate scheduling, control beamforming,
etc. so as to lower interference to mobile device(s) within cluster
902 not being served thereby. Cooperative beamforming can leverage
medium backhaul (control) requirements and can be less sensitive to
channel state information (CSI) as compared to inter-site packet
sharing. Hence, cooperative beamforming can be considered as an
alternative to inter-site packet sharing based on a performance
differential; however, the claimed subject matter is not so
limited.
[0096] Turning to FIG. 10, illustrated is an example system 1000
that effectuates cooperative silence (CS) within a cluster 1002 in
a wireless communication environment. Cluster 1002 includes base
stations 1004, 1006, and 1008 and mobile devices 1010, 1012, and
1014 (e.g., third order strategy, . . . ). As shown, base station
1004 can serve mobile device 1010, and base station 1008 can serve
mobile device 1014. Further, base station 1006 can be silent for
the benefit of mobile devices 1010 and 1014. Thus, cooperative
silence can include a node (e.g., base station 1006, . . . )
abstaining from transmission when it is beneficial for an entire
neighborhood (e.g., to remove interference, . . . ). Moreover,
cooperative silence can leverage minimum backhaul and channel state
information (CSI) requirements. It is to be appreciated, however,
that the claimed subject matter is not limited to the
foregoing.
[0097] With reference to FIG. 11, illustrated is an example system
1100 in which non-cooperative transmissions can be effectuated in a
wireless communication environment. System 1100 includes two
clusters 1102 and 1104. Cluster 1102 includes a base station 1106
and a mobile device 1108, and cluster 1104 includes a base station
1110 and a mobile device 1112. As shown, cluster 1102 and cluster
1104 each leverage first order strategies; however, it is to be
appreciated that the claimed subject matter is not so limited.
According to an illustration, when base station 1106 sends a
transmission to mobile device 1108, an impact of interference
associated with such transmission upon cluster 1104 need not be
considered (e.g., base station 1106 need not consider interference
caused to mobile device 1112, . . . ). As described herein, each
cluster 1102-1104 can dynamically change in time, and at any point
in time, cooperation technique(s) can be leveraged within each
cluster 1102-1104; yet, clusters 1102 and 1104 need not cooperate
with each other, which can result in non-cooperative
interference.
[0098] According to an example, non-cooperative interference
between clusters 1102-1104 can be treated in a similar manner as
compared to traditional base stations in conventional networks.
Thus, base station 1110 can lack knowledge or control of operations
effectuated within cluster 1102. Rather, base station 1110 can
estimate interference caused by base stations in other clusters
(e.g., base station 1106 in cluster 1102, . . . ) to mobile device
1112 without knowing beams, powers, etc. utilized by the base
stations in the other clusters. For instance, base station 1110 can
use long term information in order to schedule mobile device 1112,
and the like.
[0099] Moreover as cluster size is increased, non-cooperative
interference can decrease. For instance, when clusters are large,
an amount of cooperation can increase; yet, a tradeoff associated
with larger clusters is increased complexity (e.g., more scheduling
decisions within the cluster, more possible local strategies to
consider, . . . ). Thus, as described herein, a constraint can be
placed upon a network that controls a maximum strategy order that
can be employed (e.g., the maximum strategy order can be a second
order, a third order, a higher order, . . . ).
[0100] While an extrinsic message passed between base stations can
indicate utility and a set of constraints on a target node
associated with the utility, other details on the strategy
underlying a message typically can be unknown to the target node
unless the target node is involved in the strategy (e.g., unless
the message refers to a cooperative strategy involving the target
node, . . . ). For instance, the other details can include assumed
power spectral density (PSD), beams used by the target node, and so
forth. Hence, the target node has to assume long term interference
from the source node when evaluating its own local strategy (e.g.,
long term interference can be measured based on cell null pilots, .
. . ). Long term interference often can be sufficient as the target
node tries to avoid scheduling mobile devices when their dominant
interferers are not cooperating. Yet, accounting for dominant
interferers can be less important due to a limited impact on a
spectral efficiency of a mobile device and/or averaging across many
such interferers. Also, it can be harder to extract gains, leading
to coordinating with interferers. Moreover, more accurate
accounting for interference caused by non-cooperative strategies
can be beneficial in some scenarios since this can allow for
substantial reduction in complexity by reducing strategy order.
[0101] Thus, in addition to extrinsic utility values and a list of
involved common neighbors (e.g., base station(s) and mobile
device(s), . . . ), a source node can pass to a target node
assumptions on strategy parameters of the target node that affect
the extrinsic value. The information can be summarized, for
instance, as interference level caused by the target node to the
mobile device(s) involved in the strategy underlying that extrinsic
message. Further, the source node can define extrinsic messages
conesponding to multiple values of such parameters conesponding to
the same or different underlying strategies. Different extrinsic
messages can correspond to different values of the interference
seen from the target node to the same or different sets of mobile
devices involved in strategies underlying these extrinsic messages.
The choice of multiple messages can be driven by need to serve
mobile device(s) without cooperation from a target node which can
be a dominant interferer (e.g., source node reports messages for
different non-cooperative interference levels if mobile devices are
exposed to the same set of dominant interferences and/or when the
target node often denies cooperation, . . . ).
[0102] Turning to FIG. 12, illustrated is a system 1200 that
exchanges interference information as part of a message passing
strategy to manage non-cooperative interference in a wireless
communication environment. System 1200 includes base station 1202,
base station 1204, base station 1206, and base station 1208 (e.g.,
nodes 1202, 1204, 1206, and 1208, . . . ). Further, mobile devices
1210 and 1212 can be within a handoff region of base stations
1202-1208 in system 1200.
[0103] According to the depicted example, under a local strategy
1214, base stations 1202 and 1204 can serve mobile device 1210
(e.g., assuming a maximum strategy order is limited to 2 or 3, . .
. ). Base stations 1202-1204 involved in local strategy 1214 can
each compute its local utility under multiple assumptions on
transmit power spectral density and/or beams of every base station
not involved in local strategy 1214 (e.g., for base stations 1206
and 1208, . . . ) and formulate the corresponding multiple
extrinsic messages for the base stations not involved in local
strategy 1214. For instance, base station 1202 can consider local
strategy 1214 that serves mobile device 1210 jointly with base
station 1204. In this case, base station 1202 can evaluate local
utility of local strategy 1214 under various cases of PSD settings
and/or beam constraints by base station 1206 and 1208. Then, base
station 1202 can formulate extrinsic messages to base station 1206
and 1208 accordingly. As a function thereof, base stations 1206 and
1208 can each compute projected utilities of various local
strategies consistent with cases of the received extrinsic messages
and respective constraints on beams and power spectral density.
Thus, base stations 1206 and 1208 can have a more accurate estimate
of utilities by having knowledge of interference caused by base
stations 1202 and 1204 in a separate cluster (e.g., local strategy
1214, . . . ), which can impact clustering decisions, scheduling
decisions, and so forth.
[0104] By way of further illustration, FIGS. 13-15 illustrate
example graphs associated with a belief propagation framework for
interference avoidance and CoMP that can be implemented in
connection with the techniques described herein. While FIGS. 13-15
and the accompanying discussion below depict various examples in
the context of static clusters (e.g., each with a cluster
controller, . . . ), it is contemplated that these approaches can
be extended to clusters that are dynamically formed over time.
Hence, the below techniques can be leveraged upon dynamically
selecting an optimal set of local strategies across a network at a
given time in a distributed manner. Moreover, it is contemplated
that the distributed clustering concept noted herein can
accommodate static clusters. It is contemplated, however, that the
claimed subject matter is not limited by the below discussion.
[0105] For instance, the static clustering concept can be based on
a notion of static master clusters based on deployment and backhaul
topology. Cooperation can be possible within a master cluster and
interference management can handle boundaries. Static clustering
can be based on the Remote Radio Head (RRH) concept that can
include light remote nodes connected to a macro node via dedicated
lines (e.g., cable, fiber, . . . ), possibly with a centralized
processing architecture. Further, the remote nodes can be
independent base stations. Distributed clustering can support using
utility weights to define master clusters. For example, strategies
that stretch across RRH boundaries can be assigned zero utility
weight to prevent cooperation across such boundaries. By way of
another example, different utility weights can be used for
different strategies that stretch across RRH consistent with
inter-RRH backhaul quality (e.g., inter-site packet sharing may be
unable to be used while distributed beamforming can be used, . . .
). Pursuant to a further example, clustering across RRH boundaries
can be explicitly disabled. Static clustering can be beneficial if
RRH is a target scenario. It is to be appreciated, however, that
the claimed subject matter is not limited to the foregoing.
[0106] Consider a Radio Access Network (RAN) in a wireless cellular
system, defined by a set of base station transceivers (BTs),
denoted B={BT.sub.1, BT.sub.2, . . . }. A base station transceiver
(BT) can refer to an omni-directional cell/base station (e.g.,
node, . . . ), or a single sector of a sectorized base station.
Each base station transceiver (BT) can have one or more transmit
antennas and one or more receive antennas, used to communicate with
user terminals (UT) (e.g., mobile devices, . . . ) over a wireless
channel.
[0107] Each user terminal (UT) in the wireless cellular system can
select a serving base station transceiver based on various
criteria. The Interference Management Set (IMset) of a user
terminal can include a serving BT, along with other BTs whose
long-term forward link (FL) (reverse link (RL)) signal strength
exceeds (Q-X) dB. In this expression, the term Q can denote the
long-term FL (RL) received signal strength between the serving BT
and the user terminal (e.g., expressed in dB, . . . ), and the term
X.gtoreq.0 is an appropriately chosen parameter (e.g., 10, value
greater than 10, value less than 10, . . . ). Note that the serving
BT of a user terminal can also be part of the IMset of the user
terminal. The IMset of the user terminal extends the notion of
active set in CDMA systems, and can include potentially dominant
interferers (interferees) of the user terminal. IM.sub.u.OR right.B
can denote the IMset of the user terminal u.
[0108] A cluster can be a predefined (static) subset of BTs in the
RAN. According to another example, a cluster can be dynamically
formed as described herein. Let C={C.sub.1, C.sub.2, . . . ,
C.sub.L} denote a set of all clusters defined in the RAN (e.g., at
a given time, . . . ), wherein C.sub.j.OR right.B for each j=1, 2,
. . . , L. Pursuant to an example, different clusters can overlap
with each other (e.g., have a non-empty intersection, . . . ). By
way of further example, different clusters can be non-overlapping
(e.g., the intersection can be an empty set, . . . ). Moreover, the
clusters can be equipped with a logical entity called a cluster
controlled, which can be physically embedded in one of the BTs in
the cluster. Additionally or alternatively, functions described
below as being carried out by the cluster controlled can be
effectuated by one or more BTs in the cluster.
[0109] BTs in a cluster can be connected to the cluster controller
through low-latency (e.g., <1 msec, . . . ) signaling links. In
addition, certain BTs in a cluster can also be connected to their
cluster controller with low-latency (e.g., <1 msec, . . . ),
high-capacity (e.g., >100 MBps, . . . ) data links. Further,
certain pairs of cluster controllers can be connected to each other
through medium-latency (e.g., 10-20 msec, . . . ) signaling
links.
[0110] The cluster controller can instruct each BT in its cluster
to radiate a certain signal on the wireless channel. The signal
radiated by a BT can be a superposition (sum) of signals induced by
the controllers of clusters that include the BT. The signal induced
by a cluster at a BT can also be referred to as the signal
transmitted by the cluster from the given BT. The overall
transmitted signal of a cluster can refer to a combination (e.g.,
direct-product, . . . ) of signals induced by the cluster at BTs
belonging to that cluster.
[0111] A cluster that includes the serving BT of a user terminal
can be referred to as a serving cluster of the user terminal. Note
that although a user terminal can have a single serving BT, it can
have several serving clusters. Moreover, each serving cluster of a
user terminal can have access to a channel state and a state of
data queues/flows of the user terminal. S.sub.j can denote the set
of user terminals served by the cluster C.sub.j.
[0112] The cluster controller can decide the signal to be
transmitted by each BT in the cluster, and can also decide the data
carried by those signal resources to different user terminals
served by that cluster. The resource management (or scheduling)
decisions can be conveyed from the cluster controller to the BTs in
the cluster using the low-latency signaling links. In the case of
distributed beamforming, the cluster controller can assign certain
beam directions on certain subcarriers to different BTs, so as to
steer spatial null(s) towards user terminals that are being served
by neighboring BTs on the same set of subcarriers at the same time.
In the case of inter-site packet sharing, data associated with a
given user terminal can be transmitted/received from/at multiple
BTs in the cluster, provided high-speed data connectivity is
supported among the BTs of interest.
[0113] In a system without joint base station processing, each
cluster can coincide with a base station transceiver (BT). In a
system with intra-NodeB joint processing, each cluster can coincide
with a (e)NodeB, which can be a set of base station transceivers
(BTs) supported by a system of collocated bas band processors. Note
that the radio frequency (RF) modules/antennas of different BTs in
a (e)NodeB need not be collocated (e.g., as in the case of a Remote
Radio Head (RRH) architecture, . . . ).
[0114] If low-latency signaling links can be established between
any cluster controller and any BT, then the set of clusters C can
be defined such that the IMset of any user terminal is included in
some cluster in the RAN. In other words, for a user terminal u with
an IMset IM.sub.u, then IM.sub.u.OR right.C.sub.j for some cluster
index j. Yet, to limit complexity of cluster controllers, the
clusters can be configured to be smaller, subject to the above
IMset criterion.
[0115] FIGS. 13-15 illustrate several graphs based on topography of
UTs, BTs and clusters (e.g., cluster controllers (CCs), . . . )
characterizing a Radio Access Network (RAN).
[0116] Given a (undirected) graph G, two vertices p and q can be
said to be neighbors of each other if the graph G has an edge
between the two vertices. The two vertices p and q can be said to
be connected to each other if there is a path between them in the
graph G. The diameter of a graph can denote a maximum distance
between two vertices in the graph. The girth of a graph can denote
a length of a shortest cycle in the graph.
[0117] An interference graph of the Radio Access Network (RAN) can
be a graph G.sub.I (e.g., shown in FIG. 14, . . . ), each of whose
vertices represents a cluster. The graph G.sub.I can have an edge
between the vertex (cluster) C.sub.i and another vertex C.sub.j,
where j.noteq.i, if a UT served by cluster C.sub.i has a BT
belonging to cluster C.sub.j in its IMset (or vice versa). The
interference neighborhood of a cluster C.sub.j can refer to a set
of clusters C.sub.i such that G.sub.I has an edge between C.sub.i
and C.sub.j. The set of indices of clusters in the interference
neighborhood of C.sub.j can be denoted as N(j).
[0118] A cluster connectivity graph of the RAN can be a graph
G.sub.C (e.g., shown in FIG. 14, . . . ), each of whose vertices
represents a cluster. The graph G.sub.C can have an edge between
vertices (clusters) C.sub.i and C.sub.j if there exists (at least)
a medium latency signaling link between the clusters C.sub.i and
C.sub.j.
[0119] A resource negotiation graph G.sub.R (e.g., shown in FIG.
15, . . . ) can be a subgraph of the interference graph G.sub.I,
and the cluster connectivity graph G.sub.C. In other words, the
graph G.sub.R can have an edge between clusters C.sub.i and C.sub.j
only if the cluster C.sub.i is in the interference neighborhood of
cluster C.sub.j, and there is (at least) a medium latency signaling
link between the clusters C.sub.i and C.sub.j. Further, the
resource negotiation graph G.sub.R can be constructed so as to
minimize its diameter (e.g., mitigate long chains, . . . ) and
maximize its girth (e.g., mitigate short chains, . . . ). The set
of indices of clusters that are neighbors of C.sub.j in the
interference negotiation graph G.sub.R can be denoted by
N.sub.+(j). Hence, the set N.sub.+(j) can be a subset of the
interference neighborhood N(j). A complementary set can be denoted
as N.sub.-(j)N(j)\N.sub.-(j), which can represent the set of
cluster indices of interference neighbors of the cluster C.sub.j,
that do not have an edge to the cluster C.sub.j in the resource
negotiation graph.
[0120] Moreover, extended neighborhoods can be defined as
follows:
N.sup.e(j)N(j).orgate.{j},
N.sub.+.sup.e(j)N.sub.+(j).orgate.{j}
Further, it can follow that N.sub.-(j)=N.sup.e(j)\N.sup.e(j). The
set of clusters identified by the indices N.sup.e(j) can be
referred to as the extended interference neighborhood of the
cluster C.sub.j. Note if the resource negotiation graph has no
cycles of length 3, then for two neighboring clusters C.sub.i and
C.sub.j, it can follow that
N.sup.--.sup.e(i).andgate.N.sub.-.sup.e(j)={i, j}.
[0121] Referring to FIG. 13, illustrated is an example system 1300
with multiple user terminals (UTs), base station transceivers
(BTs), and cluster processors (e.g., cluster controllers (CCs), . .
. ). System 1300 shows example signaling/data links between cluster
controllers and base station transceivers.
[0122] The set of user terminals served by different clusters can
be given by: S.sub.1={UT.sub.1, UT.sub.2, UT.sub.3},
S.sub.2={UT.sub.3, UT.sub.4, UT.sub.5}, S.sub.3={UT.sub.4},
S.sub.4={UT.sub.7}, S.sub.5={UT.sub.8},
S.sub.6={UT.sub.7,UT.sub.10}, S.sub.7={UT.sub.5,UT.sub.10}.
[0123] Turning to FIG. 14, illustrated is an example depiction 1400
of an interference graph G.sub.I and a cluster connectivity graph
G.sub.C corresponding to system 1300 of FIG. 13. Moreover, FIG. 15
illustrates a resource negotiation graph G.sub.R corresponding to
system 1300 of FIG. 13.
[0124] In the depicted example, the cluster connectivity graph as
well as the resource negotiation graph can have two connected
components, with vertex sets {C.sub.1, C.sub.2, C.sub.3} and
{C.sub.4, C.sub.5, C.sub.6, C.sub.7}. It can also be seen that
N.sub.+(2)={3}, N.sub.-(2)={6,7}, and N.sub.G.sub.I(2)={3, 6,
7}.
[0125] The resource negotiation graph need not have edges between
pairs of interference neighbors (C.sub.2,C.sub.6),
(C.sub.2,C.sub.7) and (C.sub.3,C.sub.7) because there is no
signaling connectivity between these pairs of clusters. On the
other hand, there is no edge between interference neighbors
(C.sub.4,C.sub.7) even though they lack a signaling link there
between; this can be done so as to eliminate 3-cycles in the
resource negotiation graph G.sub.R.
[0126] A signal resource element can refer to a combination of one
(OFDM) subcarrier, one (OFDM) symbol and one spatial beam. A
spatial beam is a complex linear combination of transmit antenna
weights, or precisely, a complex-valued beamforming vector of unit
norm, each of whose components refers to a transmit antenna of a BT
in the network. A beam is said to be localized to a BT if all
non-zero components of the beamforming vector correspond to
transmit antennas of the given BT. A beam is said to be localized
to a cluster if all non-zero components of the beamforming vector
correspond to transmit antennas of some BT in the given cluster. A
signal resource block is a set of signal resource elements, all of
whose beams are localized to (at least) one cluster. Typically, a
signal resource block can be defined by a Cartesian product of a
set of (OFDM) subcarriers, a set of (OFDM) symbols/time-slots, and
a set of (spatial) beams localized to a cluster.
[0127] Recall that a cluster (controller) induces each BT in the
cluster to transmit a certain signal, and that transmit signal of a
cluster can refer to the collection (direct sum) of the signals the
cluster induces at each of its BTs. The transmit power
p(c,r).gtoreq.0 of a cluster c on a resource block r can refer to a
power of the signal obtained by (orthogonally) projecting the
transmit signal of the cluster on to the signal subspace spanned by
the resource elements in the resource block r. For instance, if the
resource block r includes a certain set of subcarriers over all
symbols, coupled with all possible beams that can be formed by a
particular BT, then the transmit power p(c,r).gtoreq.0 of the
cluster c on the resource block r refers to the total power of the
signal transmitted by the cluster from the given BT on all
subcarriers included in the resource block. In another example, if
the resource block r refers to a certain beam direction at each BT
belonging to the cluster c, then p(c,r).gtoreq.0 can refer to the
sum of the power of signals transmitted by the cluster along the
given beam direction from each BT belonging to that cluster.
[0128] Suppose the RAN defines a set of signal resource blocks
R={R.sub.1, R.sub.2, R.sub.3, . . . , R.sub.N}. A transmit power
profile of a cluster C.sub.j is a non-negative real-valued vector
P.sub.j(P.sub.j,1, P.sub.j,2, P.sub.j,3, . . . , P.sub.j,N), which
satisfies P.sub.j,k=0 unless the resource block R.sub.k is
localized to the cluster C.sub.j. This condition can capture the
fact that a cluster controller typically cannot induce any signal
at a BT not belonging to the cluster. If a cluster C.sub.j is
allocated a transmit power profile P.sub.j, then the signal
transmitted by the cluster C.sub.j satisfies the inequalities
p(C.sub.j,R.sub.k).ltoreq.P.sub.j,k for each 1.ltoreq.k.ltoreq.N.
In other words, the cluster typically does not transmit a signal
whose power exceeds the allocated profile on a resource block
R.sub.k. Two transmit power profiles P.sub.i and P.sub.j allocated
to clusters C.sub.i and C.sub.j respectively are said to be
non-overlapping if there is no component k such that both P.sub.i,k
and P.sub.j,k are positive. Note that (valid) power profiles
P.sub.i and P.sub.j of two clusters C.sub.i and C.sub.j are
non-overlapping if none of the resource blocks {, R.sub.k} are
localized to both clusters, or if the clusters C.sub.i and C.sub.j
do not share any BTs.
[0129] The transmit power profile allocated to each cluster
determines the degrees of freedom with which the cluster can serve
its user terminals. Over a given duration of time (scheduling
epoch), each cluster can be allocated a transmit power profile
based on an inter-cluster negotiation process. Once the transmit
power profile is allocated to each cluster, the cluster can manage
its resources and data transmissions so as to optimize a certain
utility function. These concepts and mechanisms are described in
more detail below.
[0130] The following relates to utility metrics that can be
employed. Suppose that each cluster C.sub.l can have a transmit
power profile P.sub.l(t) at time t. A (maximum) strength of the
signal received by a user terminal from its serving cluster C.sub.j
on each of the signal resource blocks can be determined by the
transmit power profile P.sub.j of the serving cluster C.sub.j. On
the other hand, a (maximum) interference power received by the same
user terminal on each of the signal resource blocks can be
determined by the transmit power profile P.sub.l of the clusters
C.sub.l that include a BT in the IMset of the user terminal. It
follows that the signal to interference plus noise ratio (SINR) on
a resource block that the cluster C.sub.j can achieve at each of
the user terminals can be determined by the combination of transmit
power profiles {P.sub.l |l.epsilon.N.sup.e(j)}. At each scheduling
opportunity, the cluster controller at C.sub.j can allocate signal
resources and packet formats to each of its users, which can result
in certain data rates (e.g., consistent with the SINR, . . . )
achieved by the users served by the cluster, at each time instant
t. Let r.sub.u,j denote the data rate provided to the user terminal
u.epsilon.S.sub.j by the cluster C.sub.j at any given time.
[0131] It can be evident that the set of combinations of data
rates
{ u .di-elect cons. S j r u , j } ##EQU00015##
that can be provided by the cluster C.sub.j to user terminals
served by that cluster is determined by the transmit power profiles
{P.sub.l |l.epsilon.N.sup.e(j)}. This set of achievable data rate
combinations can be denoted by
.GAMMA. j ( l .di-elect cons. N + e ( j ) P _ l ) .
##EQU00016##
[0132] Let
U j , l ( .theta. u .di-elect cons. S j r u , j ( t ) )
##EQU00017##
denote the local marginal utility metric (e.g., marginal with
respect to time, . . . ) achieved by the cluster C.sub.j at time l,
if it allocates the data rate combination
u .di-elect cons. S j r u , j ##EQU00018##
to its users at time l. For instance, under the proportional fair
scheduling of best-effort traffic, the local marginal utility
function can have the form
U j , l ( .theta. u .di-elect cons. S j r u , j ) = u .di-elect
cons. S j r u , j T u ( t ) , ##EQU00019##
where T.sub.u(t) is the average (filtered) throughput of the user u
at time l. More generally, the local, marginal utility function can
have the form
U j , l ( .theta. u .di-elect cons. S j r u , j ) = u .di-elect
cons. S j .pi. u ( l ) r u , j , where .pi. u ( l )
##EQU00020##
represents the scheduling priority of the user u at time l. The
scheduling priority .pi..sub.u,l can depend upon the transmission
deadline of packets in the data queues of the user, as well as
historical throughput provided to various data streams associated
with the user terminal.
[0133] Given the transmit power profiles
1 .ltoreq. l .ltoreq. L P _ l ##EQU00021##
of different clusters, the optimal scheduler for the cluster
C.sub.j at time l can allocate the data rate combination from the
set
.GAMMA. j ( l .di-elect cons. N + e ( j ) P _ l ) ##EQU00022##
that maximizes the local marginal utility function U.sub.j,l( . . .
). In other words, the optimal scheduling policy at the cluster
C.sub.j results in the following local marginal utility metric:
U j , l * ( .theta. l .di-elect cons. N - e ( j ) P _ l ) = .DELTA.
sup u .di-elect cons. S j r u , j .di-elect cons. .GAMMA. j ( l
.di-elect cons. N + e ( j ) P _ l ) U j , l ( .theta. u .di-elect
cons. S j r u , j ) . ##EQU00023##
[0134] The transmit power profiles {P.sub.j} can be chosen to
maximize the global marginal utility function as follows:
U l * ( .theta. 1 .ltoreq. l .ltoreq. L P _ l ) = .DELTA. l = 1 L U
j , l * ( .sym. l .di-elect cons. N + e ( j ) P _ l ) .
##EQU00024##
However, resource negotiation can be restricted to message
exchanges between neighboring clusters in the resource negotiation
graph G.sub.R, which is a subgraph of the interference graph
G.sub.I. To this end, the projected marginal utility functions can
be defined as set forth below.
[0135] For this purpose, each cluster C.sub.j can use a nominal
transmit power profile P.sub.j,m.sup.nom(l) for each cluster
C.sub.m.epsilon.N.sub.-(j). The projected, local marginal utility
function of a cluster C.sub.j can thus be given by:
U ~ j , l ( .theta. l .di-elect cons. N + e ( j ) P _ l ( l ) ) =
.DELTA. U j , l * ( .theta. l .di-elect cons. N + e ( j ) P _ l ( l
) .theta. m .di-elect cons. N - ( j ) P _ j , m nom ) = sup u
.di-elect cons. S j r u , j U j , l ( .sym. u .di-elect cons. S j r
u , j ) ##EQU00025##
The supremum with respect to the argumen
u .di-elect cons. S j r u , j ##EQU00026##
can be taken over the set
.GAMMA. j ( { .theta. l .di-elect cons. N + e ( j ) P _ l ( t ) } {
.theta. m .di-elect cons. N - ( j ) P _ j , m nom } )
##EQU00027##
of all achievable data rate combinations subject to the given
transmit profiles.
[0136] It can be desired to maximize the projected, global marginal
utility function
U ~ l ( .theta. 1 .ltoreq. l .ltoreq. L P _ l ( t ) ) = .DELTA. j =
1 L U ~ j , l ( .theta. l .di-elect cons. N - e ( j ) P _ l ( t ) )
##EQU00028##
by choosing the optimal transmit power profiles P.sub.j(t) for each
cluster C.sub.j. This can be accomplished through a message passing
algorithm described below.
[0137] Further, a resource negotiation algorithm can be supported.
Suppose clusters C.sub.p and C.sub.q have an edge between them in
the resource negotiation graph G.sub.R. The message from cluster
C.sub.p to cluster C.sub.q can include the function:
M i j ( .sym. l .di-elect cons. N + e ( i ) N - e ( j ) P _ l ( t )
) = sup { P _ l l N - e ( j ) } [ U i , l ( .theta. l .di-elect
cons. N - e ( j ) P _ l ( t ) ) + m .di-elect cons. N + ( i ) \ { j
} M m i ( .theta. l .di-elect cons. N + e ( m ) N + e ( i ) P _ l )
] ##EQU00029##
[0138] After the above message passing algorithm converges, each
cluster C.sub.i can find the transmit power profile
.theta. l .di-elect cons. N + e ( i ) P _ l = .theta. l .di-elect
cons. N + e ( i ) P ^ _ l ##EQU00030##
that maximizes the below expression:
U ~ j , l ( .theta. l .di-elect cons. N + e ( i ) P _ l ) + m
.di-elect cons. N + ( i ) M m i ( .theta. l .di-elect cons. N + e (
m ) N - e ( i ) P _ l ) = M j i ( .theta. l .di-elect cons. N - e (
i ) N - e ( j ) P _ l ( l ) ) + M j i ( .theta. l .di-elect cons. N
- e ( i ) N - e ( j ) P _ l ( l ) ) ##EQU00031##
for any neighbor C.sub.j of the cluster C.sub.i in the graph
G.sub.R. Further, the cluster C.sub.i can select the power profile
{circumflex over (P)}.sub.i for its own signal transmissions (e.g.,
through the associated BTs, . . . ).
[0139] Moreover, if the graph G.sub.R lacks 3-cycles, then
N.sub.+.sup.e(i).andgate.N.sub.-.sup.e(j)={i, j}.
[0140] Note that in the above message passing algorithm, a cluster
C.sub.i can send a message to its neighboring clusters C.sub.j for
each value of the transmit power profile vector
l .di-elect cons. N + e ( i ) N + e ( j ) P _ l ( l ) .
##EQU00032##
In order to minimize the number of messages to be exchanged, the
set of valid profiles P.sub.j of a cluster C.sub.j can be
partitioned into a small number of subsets, {Q.sub.j,1, Q.sub.j,2,
. . . , Q.sub.j,n.sub.j}.PI..sub.j. In other words,
Q.andgate.Q'=.phi. whenever Q.noteq.Q'.epsilon..PI..sub.j, and the
union of all sets Q.epsilon..PI..sub.j is the set of all power
profiles P.sub.j=.theta..sub.iP.sub.j,i such that P.sub.j,i=0
unless the resource block R.sub.i is localized to the cluster
C.sub.j. In what follows, Q.sub.j can be used to denote a generic
element of the partition .PI..sub.j (e.g.,
Q.sub.j.epsilon..PI..sub.j, . . . ). Note that Q.sub.j can
represent a subset of transmit power profiles that are localized to
the cluster C.sub.j.
[0141] Further, the quantized, projected marginal utility functions
can be defined for all Q.sub.l.epsilon..PI..sub.l as follows:
U ~ j , l q ( .theta. l .di-elect cons. N + e ( j ) Q l ) = sup P _
l .di-elect cons. Q l U ~ j , l ( .theta. l .di-elect cons. N + e (
j ) P _ l ) . ##EQU00033##
[0142] The message passing equations, thus, can take the form:
M i j q ( .sym. l .di-elect cons. N + e ( i ) N - e ( j ) Q l ) =
sup Q l .di-elect cons. .PI. l : l N + e ( j ) [ U ~ i , l (
.theta. l .di-elect cons. N - e ( j ) Q l ) + m .di-elect cons. N +
( i ) \ { j } M m i q ( .theta. l .di-elect cons. N + e ( m ) N + e
( i ) Q l ) ] ##EQU00034##
[0143] After the above message passing algorithm converges, each
cluster C.sub.i can determine the transmit power profile subset
.theta. l .di-elect cons. N + e ( i ) Q l = .theta. l .di-elect
cons. N + e ( i ) Q ^ l ##EQU00035##
that maximizes the expression
U ~ i , l q ( .theta. l .di-elect cons. N - e ( i ) Q l ) + m
.di-elect cons. N + ( i ) M m i q ( .theta. l .di-elect cons. N + e
( m ) N + e ( i ) Q l ) = M j i q ( .theta. l .di-elect cons. N + e
( j ) N + e ( i ) Q l ) + M i j q ( .theta. l .di-elect cons. N + e
( i ) N - e ( j ) Q l ) ##EQU00036##
for any j.epsilon.N.sub.-(i). The foregoing can be maximized among
all subsets Q.sub.l included in the partition .PI..sub.l, for
l.epsilon.N.sub.-.sup.e(i).
[0144] Once the algorithm converges on the preferred power profile
subsets {{circumflex over (Q)}.sub.l}, these sets can be further
partitioned, and the message passing algorithm can be repeated on
the elements of the new partition. This process of successive
refinement can be continued until the optimal power profiles
{{circumflex over (P)}.sub.l} are represented with sufficient
precision by the partitioned power profile subsets {{circumflex
over (Q)}.sub.l}.
[0145] Referring to FIGS. 16-17, methodologies relating to
dynamically selecting clustering strategies in a distributed manner
in a wireless communication environment are illustrated. While, for
purposes of simplicity of explanation, the methodologies are shown
and described as a series of acts, it is to be understood and
appreciated that the methodologies are not limited by the order of
acts, as some acts may, in accordance with one or more embodiments,
occur in different orders and/or concurrently with other acts from
that shown and described herein. For example, those skilled in the
art will understand and appreciate that a methodology could
alternatively be represented as a series of interrelated states or
events, such as in a state diagram. Moreover, not all illustrated
acts may be required to implement a methodology in accordance with
one or more embodiments.
[0146] Turning to FIG. 16, illustrated is a methodology 1600 that
facilitates dynamically forming clusters in a wireless
communication environment. At 1602, local utilities of possible
local strategies involving a base station at a given time can be
evaluated. Each of the possible local strategies can include a set
of base stations (e.g., including the base station evaluating the
local utilities and possibly one or more neighboring base stations,
. . . ), a set of mobile devices (e.g., one or more mobile devices,
. . . ) served by the set of base stations, and underlying antenna
weights and power spectral densities for the base station(s) in the
set of base stations to serve the set of mobile devices. Moreover,
the possible local strategies can be subject to a limited maximum
order constraint (e.g., a maximum strategy order can be two, three,
an integer greater than three, . . . ). The local utilities, for
instance, can be summations of weighted rates that can be achieved
by at least one mobile device respectively served under each of the
possible local strategies.
[0147] At 1604, strategy and utility information can be exchanged
with at least one neighbor base station through message passing.
According to an example, message passing can be iterative; however,
the claimed subject matter is not so limited. Further, the base
station can send the strategy and utility information yielded by
the base station to the at least one neighbor base station and
receive the strategy and utility information respectively yielded
by each of the at least one neighbor base station from the at least
one neighbor base station.
[0148] By way of example, the strategy and utility information can
include a cooperative utility value and a non-cooperative utility
value. The cooperative utility value can reflect an estimate of
total utility assuming cooperation between a source (e.g., source
of the strategy and utility information, . . . ) and a target
(e.g., target of the strategy and utility information, . . . ).
Further, the non-cooperative utility value can reflect an estimate
of total utility assuming lack of cooperation between the source
and the target. According to another example, the strategy and
utility information can include a plurality of utility values
assuming various constraints upon the target, where the assumed
constraints are reported from the source to the target.
[0149] At 1606, network-wide utility estimates can be generated for
the possible local strategies as a function of the strategy and
utility information received from the at least one neighbor base
station through message passing and the evaluated local utilities.
Message passing can enable each base station to compute
network-wide utility estimates associated with respective possible
local strategies. Moreover, the network-wide utility estimates can
be further generated at least in part as a function of
non-cooperative interference information received from the at least
one neighbor base station.
[0150] At 1608, a particular local strategy from the possible local
strategies can be selected for use by the base station based upon
the network-wide utility estimates. The particular local strategy
can be selected by the base station; similarly, disparate base
stations in a wireless communication environment can each likewise
select a respective particular local strategy for use thereby. For
example, the particular local strategy can yield a maximum (e.g.,
optimal, . . . ) network-wide utility estimate as compared to
network-wide utility estimates corresponding to the remaining
possible local strategies. Moreover, the selected particular local
strategy can be non-contradictory to particular local strategies
respectively selected by disparate base stations in the network
(e.g., wireless communication environment, . . . ). Thus, clusters
dynamically formed based upon the particular local strategies
respectively selected by the base station and the disparate base
stations within the network can be non-overlapping.
[0151] Referring to FIG. 17, illustrated is a methodology 1700 that
facilitates leveraging cooperation between base stations in a
wireless communication environment. At 1702, a particular local
strategy that includes a base station can be selected as a function
of time based upon network-wide utility estimates respectively
conditioned upon implementation of the particular local strategy
and disparate possible local strategies that include the base
station. The particular local strategy can be selected by the base
station (e.g., selection can be effectuated in a distributed
manner, . . . ). Moreover, a cluster including the base station can
be dynamically formed based upon the selected particular local
strategy. At 1704, operation within a cluster formed according to
the selected particular local strategy can be coordinated. For
instance, packets can be shared amongst base stations in the
cluster (e.g., for transmission to served mobile device(s), . . .
). Moreover, scheduling within the cluster can be effectuated.
According to further examples, at least one of inter-site packet
sharing, cooperative beamforming, or cooperative silence can be
implemented within the cluster.
[0152] According to another example, the base station can exchange
transmission information (e.g., related to beams, power spectral
densities (PSDs), . . . ) with at least one base station included
in at least one different cluster (e.g., at least one different
strategy that does not include the base station, . . . ). Following
this example, the base station can assess inter-cluster
interference based upon transmission information received from the
at least one different cluster (e.g., the inter-cluster
interference assessment based upon the exchanged interference
information can be more refined compared to a long-term
interference estimate, . . . ). Moreover, the inter-cluster
interference assessment can be factored into a utility computation.
However, it is contemplated that the claimed subject matter is not
limited to the foregoing example.
[0153] It will be appreciated that, in accordance with one or more
aspects described herein, inferences can be made regarding
dynamically forming clusters in a distributed fashion in a wireless
communication environment. As used herein, the term to "infer" or
"inference" refers generally to the process of reasoning about or
inferring states of the system, environment, and/or user from a set
of observations as captured via events and/or data. Inference can
be employed to identify a specific context or action, or can
generate a probability distribution over states, for example. The
inference can be probabilistic--that is, the computation of a
probability distribution over states of interest based on a
consideration of data and events. Inference can also refer to
techniques employed for composing higher-level events from a set of
events and/or data. Such inference results in the construction of
new events or actions from a set of observed events and/or stored
event data, whether or not the events are correlated in close
temporal proximity, and whether the events and data come from one
or several event and data sources.
[0154] According to an example, one or more methods presented above
can include making inferences pertaining to determining
network-wide utilities associated with differing possible local
strategies. By way of further illustration, an inference can be
made related to identifying constraints associated with various
possible local strategies. It will be appreciated that the
foregoing examples are illustrative in nature and are not intended
to limit the number of inferences that can be made or the manner in
which such inferences are made in conjunction with the various
embodiments and/or methods described herein.
[0155] FIG. 18 is an illustration of a mobile device 1800 that can
be employed in connection with various aspects described herein.
Mobile device 1800 comprises a receiver 1802 that receives a signal
from, for instance, a receive antenna (not shown), and performs
typical actions thereon (e.g., filters, amplifies, downconverts,
etc.) the received signal and digitizes the conditioned signal to
obtain samples. Receiver 1802 can be, for example, an MMSE
receiver, and can comprise a demodulator 1804 that can demodulate
received symbols and provide them to a processor 1806 for channel
estimation. Processor 1806 can be a processor dedicated to
analyzing information received by receiver 1802 and/or generating
information for transmission by a transmitter 1812, a processor
that controls one or more components of mobile device 1800, and/or
a processor that both analyzes information received by receiver
1802, generates information for transmission by transmitter 1812,
and controls one or more components of mobile device 1800.
[0156] Mobile device 1800 can additionally comprise memory 1808
that is operatively coupled to processor 1806 and that can store
data to be transmitted, received data, and any other suitable
information related to performing the various actions and functions
set forth herein.
[0157] It will be appreciated that the data store (e.g., memory
1808) described herein can be either volatile memory or nonvolatile
memory, or can include both volatile and nonvolatile memory. By way
of illustration, and not limitation, nonvolatile memory can include
read only memory (ROM), programmable ROM (PROM), electrically
programmable ROM (EPROM), electrically erasable PROM (EEPROM), or
flash memory. Volatile memory can include random access memory
(RAM), which acts as external cache memory. By way of illustration
and not limitation, RAM is available in many forms such as
synchronous RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM
(SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM
(ESDRAM), Synchlink DRAM (SLDRAM), and direct Rambus RAM (DRRAM).
The memory 1808 of the subject systems and methods is intended to
comprise, without being limited to, these and any other suitable
types of memory.
[0158] Mobile device 1800 still further comprises a modulator 1810
and a transmitter 1812 that transmits data, signals, etc. to a base
station. Although depicted as being separate from the processor
1806, it is to be appreciated that modulator 1810 can be part of
processor 1806 or a number of processors (not shown).
[0159] FIG. 19 is an illustration of a system 1900 that dynamically
selects a local strategy to employ over time in a wireless
communication environment. System 1900 comprises a base station
1902 (e.g., access point, . . . ) with a receiver 1910 that
receives signal(s) from one or more mobile devices 1904 through a
plurality of receive antennas 1906, and a transmitter 1924 that
transmits to the one or more mobile devices 1904 through a transmit
antenna 1908. Moreover, base station 1902 can receive signal(s)
with receiver 1910 from one or more disparate base stations through
the plurality of receive antennas 1906 and/or transmit to one or
more disparate base stations with transmitter 1924 through the
transmit antenna 1908. According to another illustration, base
station 1902 can receive signal(s) from (e.g., with receiver 1910,
. . . ) and/or transmit signal(s) to (e.g., with transmitter 1924,
. . . ) one or more disparate base stations via a backhaul.
Receiver 1910 can receive information from receive antennas 1906
and is operatively associated with a demodulator 1912 that
demodulates received information. Demodulated symbols are analyzed
by a processor 1914 that can be similar to the processor described
above with regard to FIG. 18, and which is coupled to a memory 1916
that stores data to be transmitted to or received from mobile
device(s) 1904 and/or disparate base station(s) and/or any other
suitable information related to performing the various actions and
functions set forth herein. Processor 1914 is further coupled to a
metric evaluation component 1918 and/or a negotiation component
1920. Metric evaluation component 1918 can be substantially similar
to metric evaluation component 408 of FIG. 4 and/or negotiation
component 1920 can be substantially similar to negotiation
component 410 of FIG. 4. Metric evaluation component 1918 can
analyze local utilities associated with possible local strategies
that can cover base station 1902. Moreover, negotiation component
1920 can effectuate message passing to exchange strategy and
utility information between base station 1902 and neighbor base
stations. Further, received strategy and utility information can be
evaluated by metric evaluation component 1918 to generate
network-wide utility estimates conditioned upon each of the
possible local strategies. Based upon such network-wide utility
estimates, base station 1902 can elect a particular one of the
possible local strategies. Moreover, although not shown, it is to
be appreciated that base station 1902 can further include a
clustering component (e.g., substantially similar to clustering
component 406 of FIG. 4, . . . ) and/or a cooperation component
(e.g., substantially similar to cooperation component 710 of FIG.
7, . . . ). Base station 1902 can further include a modulator 1922.
Modulator 1922 can multiplex a frame for transmission by a
transmitter 1924 through antennas 1908 to mobile device(s) 1904 in
accordance with the aforementioned description. Although depicted
as being separate from the processor 1914, it is to be appreciated
that metric evaluation component 1918, negotiation component 1920,
and/or modulator 1922 can be part of processor 1914 or a number of
processors (not shown).
[0160] FIG. 20 shows an example wireless communication system 2000.
The wireless communication system 2000 depicts one base station
2010 and one mobile device 2050 for sake of brevity. However, it is
to be appreciated that system 2000 can include more than one base
station and/or more than one mobile device, wherein additional base
stations and/or mobile devices can be substantially similar or
different from example base station 2010 and mobile device 2050
described below. In addition, it is to be appreciated that base
station 2010 and/or mobile device 2050 can employ the systems
(FIGS. 1-12, 18-19 and 21) and/or methods (FIGS. 16-17) described
herein to facilitate wireless communication there between.
[0161] At base station 2010, traffic data for a number of data
streams is provided from a data source 2012 to a transmit (TX) data
processor 2014. According to an example, each data stream can be
transmitted over a respective antenna. TX data processor 2014
formats, codes, and interleaves the traffic data stream based on a
particular coding scheme selected for that data stream to provide
coded data.
[0162] The coded data for each data stream can be multiplexed with
pilot data using orthogonal frequency division multiplexing (OFDM)
techniques. Additionally or alternatively, the pilot symbols can be
frequency division multiplexed (FDM), time division multiplexed
(TDM), or code division multiplexed (CDM). The pilot data is
typically a known data pattern that is processed in a known manner
and can be used at mobile device 2050 to estimate channel response.
The multiplexed pilot and coded data for each data stream can be
modulated (e.g., symbol mapped) based on a particular modulation
scheme (e.g., binary phase-shift keying (BPSK), quadrature
phase-shift keying (QPSK), M-phase-shift keying (M-PSK),
M-quadrature amplitude modulation (M-QAM), etc.) selected for that
data stream to provide modulation symbols. The data rate, coding,
and modulation for each data stream can be determined by
instructions performed or provided by processor 2030. Memory 2032
can store program code, data, and other information used by
processor 2030 or other components of base station 2010.
[0163] The modulation symbols for the data streams can be provided
to a TX MIMO processor 2020, which can further process the
modulation symbols (e.g., for OFDM). TX MIMO processor 2020 then
provides N.sub.T modulation symbol streams to N.sub.T transmitters
(TMTR) 2022a through 2022t. In various embodiments, TX MIMO
processor 2020 applies beamforming weights to the symbols of the
data streams and to the antenna from which the symbol is being
transmitted.
[0164] Each transmitter 2022 receives and processes a respective
symbol stream to provide one or more analog signals, and further
conditions (e.g., amplifies, filters, and upconverts) the analog
signals to provide a modulated signal suitable for transmission
over the MIMO channel. Further, N.sub.T modulated signals from
transmitters 2022a through 2022t are transmitted from N.sub.T
antennas 2024a through 2024t, respectively.
[0165] At mobile device 2050, the transmitted modulated signals are
received by N.sub.R antennas 2052a through 2052r and the received
signal from each antenna 2052 is provided to a respective receiver
(RCVR) 2054a through 2054r. Each receiver 2054 conditions (e.g.,
filters, amplifies, and downconverts) a respective signal,
digitizes the conditioned signal to provide samples, and further
processes the samples to provide a corresponding "received" symbol
stream.
[0166] An RX data processor 2060 can receive and process the
N.sub.R received symbol streams from N.sub.R receivers 2054 based
on a particular receiver processing technique to provide N.sub.T
"detected" symbol streams. RX data processor 2060 can demodulate,
deinterleave, and decode each detected symbol stream to recover the
traffic data for the data stream. The processing by RX data
processor 2060 is complementary to that performed by TX MIMO
processor 2020 and TX data processor 2014 at base station 2010.
[0167] A processor 2070 can periodically determine which precoding
matrix to utilize as discussed above. Further, processor 2070 can
formulate a reverse link message comprising a matrix index portion
and a rank value portion.
[0168] The reverse link message can comprise various types of
information regarding the communication link and/or the received
data stream. The reverse link message can be processed by a TX data
processor 2038, which also receives traffic data for a number of
data streams from a data source 2036, modulated by a modulator
2080, conditioned by transmitters 2054a through 2054r, and
transmitted back to base station 2010.
[0169] At base station 2010, the modulated signals from mobile
device 2050 are received by antennas 2024, conditioned by receivers
2022, demodulated by a demodulator 2040, and processed by a RX data
processor 2042 to extract the reverse link message transmitted by
mobile device 2050. Further, processor 2030 can process the
extracted message to determine which precoding matrix to use for
determining the beamforming weights.
[0170] Processors 2030 and 2070 can direct (e.g., control,
coordinate, manage, etc.) operation at base station 2010 and mobile
device 2050, respectively. Respective processors 2030 and 2070 can
be associated with memory 2032 and 2072 that store program codes
and data. Processors 2030 and 2070 can also perform computations to
derive frequency and impulse response estimates for the uplink and
downlink, respectively.
[0171] It is to be understood that the embodiments described herein
can be implemented in hardware, software, firmware, middleware,
microcode, or any combination thereof. For a hardware
implementation, the processing units can be implemented within one
or more application specific integrated circuits (ASICs), digital
signal processors (DSPs), digital signal processing devices
(DSPDs), programmable logic devices (PLDs), field programmable gate
arrays (FPGAs), processors, controllers, micro-controllers,
microprocessors, other electronic units designed to perform the
functions described herein, or a combination thereof.
[0172] When the embodiments are implemented in software, firmware,
middleware or microcode, program code or code segments, they can be
stored in a machine-readable medium, such as a storage component. A
code segment can represent a procedure, a function, a subprogram, a
program, a routine, a subroutine, a module, a software package, a
class, or any combination of instructions, data structures, or
program statements. A code segment can be coupled to another code
segment or a hardware circuit by passing and/or receiving
information, data, arguments, parameters, or memory contents.
Information, arguments, parameters, data, etc. can be passed,
forwarded, or transmitted using any suitable means including memory
sharing, message passing, token passing, network transmission,
etc.
[0173] For a software implementation, the techniques described
herein can be implemented with modules (e.g., procedures,
functions, and so on) that perform the functions described herein.
The software codes can be stored in memory units and executed by
processors. The memory unit can be implemented within the processor
or external to the processor, in which case it can be
communicatively coupled to the processor via various means as is
known in the art.
[0174] With reference to FIG. 21, illustrated is a system 2100 that
enables employing dynamically defined clusters in a wireless
communication environment. For example, system 2100 can reside at
least partially within a base station. It is to be appreciated that
system 2100 is represented as including functional blocks, which
can be functional blocks that represent functions implemented by a
processor, software, or combination thereof (e.g., firmware).
System 2100 includes a logical grouping 2102 of electrical
components that can act in conjunction. For instance, logical
grouping 2102 can include an electrical component for choosing a
particular local strategy as a function of time based upon
network-wide utility estimates respectively conditioned upon the
particular local strategy and disparate possible local strategies
2104. Moreover, logical grouping 2102 can include an electrical
component for controlling operation within a cluster dynamically
formed based upon the chosen particular local strategy 2106.
Further, logical grouping 2102 can optionally include an electrical
component for exchanging information utilized to evaluate the
network-wide utility estimates with at least one neighbor base
station 2108. Additionally, system 2100 can include a memory 2110
that retains instructions for executing functions associated with
electrical components 2104, 2106, and 2108. While shown as being
external to memory 2110, it is to be understood that one or more of
electrical components 2104, 2106, and 2108 can exist within memory
2110.
[0175] The various illustrative logics, logical blocks, modules,
and circuits described in connection with the embodiments disclosed
herein can be implemented or performed with a general purpose
processor, a digital signal processor (DSP), an application
specific integrated circuit (ASIC), a field programmable gate array
(FPGA) or other programmable logic device, discrete gate or
transistor logic, discrete hardware components, or any combination
thereof designed to perform the functions described herein. A
general-purpose processor can be a microprocessor, but, in the
alternative, the processor can be any conventional processor,
controller, microcontroller, or state machine. A processor can also
be implemented as a combination of computing devices, e.g., a
combination of a DSP and a microprocessor, a plurality of
microprocessors, one or more microprocessors in conjunction with a
DSP core, or any other such configuration. Additionally, at least
one processor can comprise one or more modules operable to perform
one or more of the steps and/or actions described above.
[0176] Further, the steps and/or actions of a method or algorithm
described in connection with the aspects disclosed herein can be
embodied directly in hardware, in a software module executed by a
processor, or in a combination of the two. A software module can
reside in RAM memory, flash memory, ROM memory, EPROM memory,
EEPROM memory, registers, a hard disk, a removable disk, a CD-ROM,
or any other form of storage medium known in the art. An exemplary
storage medium can be coupled to the processor, such that the
processor can read information from, and write information to, the
storage medium. In the alternative, the storage medium can be
integral to the processor. Further, in some aspects, the processor
and the storage medium can reside in an ASIC. Additionally, the
ASIC can reside in a user terminal. In the alternative, the
processor and the storage medium can reside as discrete components
in a user terminal. Additionally, in some aspects, the steps and/or
actions of a method or algorithm can reside as one or any
combination or set of codes and/or instructions on a machine
readable medium and/or computer readable medium, which can be
incorporated into a computer program product.
[0177] In one or more aspects, the functions described can be
implemented in hardware, software, firmware, or any combination
thereof. If implemented in software, the functions can be stored or
transmitted as one or more instructions or code on a
computer-readable medium. Computer-readable media includes both
computer storage media and communication media including any medium
that facilitates transfer of a computer program from one place to
another. A storage medium can be any available media that can be
accessed by a computer. By way of example, and not limitation, such
computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or
other optical disk storage, magnetic disk storage or other magnetic
storage devices, or any other medium that can be used to carry or
store desired program code in the form of instructions or data
structures and that can be accessed by a computer. Also, any
connection can be termed a computer-readable medium. For example,
if software is transmitted from a website, server, or other remote
source using a coaxial cable, fiber optic cable, twisted pair,
digital subscriber line (DSL), or wireless technologies such as
infrared, radio, and microwave, then the coaxial cable, fiber optic
cable, twisted pair, DSL, or wireless technologies such as
infrared, radio, and microwave are included in the definition of
medium. Disk and disc, as used herein, includes compact disc (CD),
laser disc, optical disc, digital versatile disc (DVD), floppy disk
and blu-ray disc where disks usually reproduce data magnetically,
while discs usually reproduce data optically with lasers.
Combinations of the above should also be included within the scope
of computer-readable media.
[0178] While the foregoing disclosure discusses illustrative
aspects and/or embodiments, it should be noted that various changes
and modifications could be made herein without departing from the
scope of the described aspects and/or embodiments as defined by the
appended claims. Furthermore, although elements of the described
aspects and/or embodiments can be described or claimed in the
singular, the plural is contemplated unless limitation to the
singular is explicitly stated. Additionally, all or a portion of
any aspect and/or embodiment can be utilized with all or a portion
of any other aspect and/or embodiment, unless stated otherwise.
* * * * *