U.S. patent application number 11/536500 was filed with the patent office on 2007-03-29 for method and apparatus for mitigating multiuser access interference.
This patent application is currently assigned to LG Electronics Inc.. Invention is credited to Shu Wang.
Application Number | 20070072550 11/536500 |
Document ID | / |
Family ID | 37900177 |
Filed Date | 2007-03-29 |
United States Patent
Application |
20070072550 |
Kind Code |
A1 |
Wang; Shu |
March 29, 2007 |
METHOD AND APPARATUS FOR MITIGATING MULTIUSER ACCESS
INTERFERENCE
Abstract
A method of mitigating interference in a multiple access
interference wireless communication system is disclosed. More
specifically, the method comprises receiving at least one signal
from at least one transmitting end, constructing a group of first
signal signatures based on a predetermined number of received
signals and at least one second signal signature, estimating a set
of values using the constructed group of first signal signatures,
and detecting desired information using the detected set of
values.
Inventors: |
Wang; Shu; (San Diego,
CA) |
Correspondence
Address: |
LEE, HONG, DEGERMAN, KANG & SCHMADEKA
801 S. FIGUEROA STREET
12TH FLOOR
LOS ANGELES
CA
90017
US
|
Assignee: |
LG Electronics Inc.
|
Family ID: |
37900177 |
Appl. No.: |
11/536500 |
Filed: |
September 28, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60722018 |
Sep 28, 2005 |
|
|
|
Current U.S.
Class: |
455/63.1 ;
375/E1.025; 455/501 |
Current CPC
Class: |
H04L 25/022 20130101;
H04J 13/00 20130101; H04B 2201/70701 20130101; H04L 25/0226
20130101; H04B 1/712 20130101; H04B 1/7105 20130101; H04B
2201/709727 20130101 |
Class at
Publication: |
455/063.1 ;
455/501 |
International
Class: |
H04B 1/00 20060101
H04B001/00; H04B 15/00 20060101 H04B015/00 |
Claims
1. A method of mitigating interference in a multiple access
interference wireless communication system, the method comprising:
receiving at least one signal from at least one transmitting end;
constructing a group of first signal signatures based on a
predetermined number of received signals and at least one second
signal signature; estimating a set of values using the constructed
group of first signal signatures; and detecting desired information
using the detected set of values.
2. The method of claim 1, wherein the group of first signal
signatures is referred to as a blind spreading matrix.
3. The method of claim 1, wherein the at least one second signal
signature is referred to as spreading sequence.
4. The method of claim 3, wherein the at least one second signal
structure is known by a receiving end.
5. The method of claim 1, wherein the set of values is referred to
as a detection vector.
6. The method of claim 1, wherein at least one of the set of values
and the desired information is acquired by any one of least squares
(LS), total least squares (TLS), mixed LS/TLS, best linear unbiased
(BLU), and minimum mean squared error (MMSE) detection schemes.
7. An apparatus for mitigating interference in a multiple access
interference wireless communication system, the system comprising:
a receiving unit for receiving at least one signal from at least
one transmitting end; a constructing unit for a group of first
signal signatures based on a predetermined number of received
signals and at least one second signal signature; an estimating
unit estimating a set of values using the constructed group of
first signal signatures; a combining unit for combining the at
least one estimated variable component with the constructed data
matrix; and a detecting unit for detecting desired information
using the detected set of values.
8. The apparatus of claim 7, wherein the group of first signal
signatures is referred to as a blind spreading matrix.
9. The apparatus of claim 8, wherein the at least one second signal
signature is referred to as spreading sequence.
10. The apparatus of claim 7, wherein the at least one second
signal structure is known by a receiving end.
11. The appartus of claim 7, wherein the set of values is referred
to as a detection vector.
12. The method of claim 7, wherein at least one of the set of
values and the desired information is acquired by any one of least
squares (LS), total least squares (TLS), mixed LS/TLS, best linear
unbiased (BLU), and minimum mean squared error (MMSE) detection
schemes.
13. A method of mitigating interference in a multiple access
interference wireless communication system, the method comprising:
receiving signals from at least one transmitting end; constructing
a blind spreading sequence matrix in which each interference is
treated as known signal and data matrix; estimating at least one
variable component by using the at least one received signals and
by using the constructed blind spreading sequence matrix;
multiplexing the at least one estimated variable component with the
constructed data matrix; and detecting data using the at least one
combined variable component
Description
[0001] This application claims the benefit of U.S. Provisional
Application No. 60/722,018, filed on Sep. 28, 2005, which is hereby
incorporated by reference.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] The present invention relates to a method for mitigating
interference, and more particularly, to a method and an apparatus
for mitigating multiuser access interference.
[0004] 2. Discussion of the Related Art
[0005] In the world of cellular telecommunications, those skilled
in the art often use the terms 1G, 2G, and 3G. The terms refer to
the generation of the cellular technology used. 1G refers to the
first generation, 2G to the second generation, and 3G to the third
generation.
[0006] 1G refers to the analog phone system, known as an AMPS
(Advanced Mobile Phone Service) phone systems. 2G is commonly used
to refer to the digital cellular systems that are prevalent
throughout the world, and include CDMAOne, Global System for Mobile
communications (GSM), and Time Division Multiple Access (TDMA). 2G
systems can support a greater number of users in a dense area than
can 1G systems.
[0007] 3G commonly refers to the digital cellular systems currently
being deployed. These 3G communication systems are conceptually
similar to each other with some significant differences.
[0008] With respect to detecting signals transmitted by single to
multi-users, typically, signals that originate from closer distance
have stronger signal than the signals transmitted from a far
distance. Moreover, stronger signals are less susceptive to
problems associated with interference.
[0009] To combat problems associated with the signals far from to
the source as opposed to near from the source, several existing
approaches or views exist, such as interference estimation and
cancellation and subspace-based approach. For example, the
interference estimation and cancellation approach reconstructs
signal interference based on channel estimation (e.g., timing,
amplitude, and phase), and thereafter, eliminates the interference
before performing detection. As another example, the subspace-based
approach estimates interfering signal subspace and constructs
multi-user detectors before performing detection.
[0010] Although these approaches work, it can be a bit complex,
involves complicated matrix subspace decomposition, and need to
assume that there are unlimited or enough received signals.
SUMMARY OF THE INVENTION
[0011] Accordingly, the present invention is directed to a method
and an apparatus for mitigating multiuser access interference that
substantially obviates one or more problems due to limitations and
disadvantages of the related art.
[0012] An object of the present invention is to provide a method of
mitigating interference in a multiple access interference wireless
communication system.
[0013] Another object of the present invention is to provide an
apparatus for mitigating interference in a multiple access
interference wireless communication system.
[0014] Additional advantages, objects, and features of the
invention will be set forth in part in the description which
follows and in part will become apparent to those having ordinary
skill in the art upon examination of the following or may be
learned from practice of the invention. The objectives and other
advantages of the invention may be realized and attained by the
structure particularly pointed out in the written description and
claims hereof as well as the appended drawings.
[0015] To achieve these objects and other advantages and in
accordance with the purpose of the invention, as embodied and
broadly described herein, a method of mitigating interference in a
multiple access interference wireless communication system includes
receiving at least one signal from at least one transmitting end,
constructing a group of first signal signatures based on a
predetermined number of received signals and at least one second
signal signature, estimating a set of values using the constructed
group of first signal signatures, and detecting desired information
using the detected set of values.
[0016] In another aspect of the present invention, an apparatus for
mitigating interference in a multiple access interference wireless
communication system includes a receiving unit for receiving at
least one signal from at least one transmitting end, a constructing
unit for a group of first signal signatures based on a
predetermined number of received signals and at least one second
signal signature, an estimating unit estimating a set of values
using the constructed group of first signal signatures, a combining
unit for combining the at least one estimated variable component
with the constructed data matrix, and a detecting unit for
detecting desired information using the detected set of values.
[0017] In a further aspect of the present invention, an apparatus
for mitigating interference in a multiple access interference
wireless communication system includes receiving signals from at
least one transmitting end, constructing a blind spreading sequence
matrix in which each interference is treated as known signal and
data matrix, estimating at least one variable component by using
the at least one received signals and by using the constructed
blind spreading sequence matrix, multiplexing the at least one
estimated variable component with the constructed data matrix, and
detecting data using the at least one combined variable
component.
[0018] It is to be understood that both the foregoing general
description and the following detailed description of the present
invention are exemplary and explanatory and are intended to provide
further explanation of the invention as claimed.
BRIEF DESCRIPTION OF THE DRAWINGS
[0019] The accompanying drawings, which are included to provide a
further understanding of the invention and are incorporated in and
constitute a part of this application, illustrate embodiment(s) of
the invention and together with the description serve to explain
the principle of the invention. In the drawings;
[0020] FIG. 1 illustrates wireless communication network
architecture;
[0021] FIG. 2A illustrates a CDMA spreading and de-spreading
process;
[0022] FIG. 2B illustrates a CDMA spreading and de-spreading
process using multiple spreading sequences;
[0023] FIG. 3 illustrates a data link protocol architecture layer
for a cdma2000 wireless network;
[0024] FIG. 4 illustrates cdma2000 call processing;
[0025] FIG. 5 illustrates the cdma2000 initialization state;
[0026] FIG. 6 illustrates the cdma2000 system access state;
[0027] FIG. 7 illustrates a conventional cdma2000 access
attempt;
[0028] FIG. 8 illustrates a conventional cdma2000 access
sub-attempt;
[0029] FIG. illustrates the conventional cdma2000 system access
state using slot offset;
[0030] FIG. 10 illustrates a comparison of cdma2000 for 1x and
1xEV-DO;
[0031] FIG. 11 illustrates a network architecture layer for a
1xEV-DO wireless network;
[0032] FIG. 12 illustrates 1xEV-DO default protocol
architecture;
[0033] FIG. 13 illustrates 1xEV-DO non-default protocol
architecture;
[0034] FIG. 14 illustrates 1xEV-DO session establishment;
[0035] FIG. 15 illustrates 1xEV-DO connection layer protocols;
and
[0036] FIG. 16 illustrates an exemplary diagram of a receiver
design.
DETAILED DESCRIPTION OF THE INVENTION
[0037] Reference will now be made in detail to the preferred
embodiments of the present invention, examples of which are
illustrated in the accompanying drawings. Wherever possible, the
same reference numbers will be used throughout the drawings to
refer to the same or like parts.
[0038] Referring to FIG. 1, a wireless communication network
architecturel is illustrated. A subscriber uses a mobile station
(MS) 2 to access network services. The MS 2 may be a portable
communications unit, such as a hand-held cellular phone, a
communication unit installed in a vehicle, or a fixed-location
communications unit.
[0039] The electromagnetic waves for the MS 2 are transmitted by
the Base Transceiver System (BTS) 3 also known as node B. The BTS 3
consists of radio devices such as antennas and equipment for
transmitting and receiving radio waves. The BS 6 Controller (BSC) 4
receives the transmissions from one or more BTS's. The BSC 4
provides control and management of the radio transmissions from
each BTS 3 by exchanging messages with the BTS and the Mobile
Switching Center (MSC) 5 or Internal IP Network. The BTS's 3 and
BSC 4 are part of the BS 6 (BS) 6.
[0040] The BS 6 exchanges messages with and transmits data to a
Circuit Switched Core Network (CSCN) 7 and Packet Switched Core
Network (PSCN) 8. The CSCN 7 provides traditional voice
communications and the PSCN 8 provides Internet applications and
multimedia services.
[0041] The Mobile Switching Center (MSC) 5 portion of the CSCN 7
provides switching for traditional voice communications to and from
a MS 2 and may store information to support these capabilities. The
MSC 2 may be connected to one of more BS's 6 as well as other
public networks, for example a Public Switched Telephone Network
(PSTN) (not shown) or Integrated Services Digital Network (ISDN)
(not shown). A Visitor Location Register (VLR) 9 is used to
retrieve information for handling voice communications to or from a
visiting subscriber. The VLR 9 may be within the MSC 5 and may
serve more than one MSC.
[0042] A user identity is assigned to the Home Location Register
(HLR) 10 of the CSCN 7 for record purposes such as subscriber
information, for example Electronic Serial Number (ESN), Mobile
Directory Number (MDR), Profile Information, Current Location, and
Authentication Period. The Authentication Center (AC) 11 manages
authentication information related to the MS 2. The AC 11 may be
within the HLR 10 and may serve more than one HLR. The interface
between the MSC 5 and the HLR/AC 10, 11 is an IS-41 standard
interface 18.
[0043] The Packet data Serving Node (PDSN) 12 portion of the PSCN 8
provides routing for packet data traffic to and from MS 2. The PDSN
12 establishes, maintains, and terminates link layer sessions to
the MS 2's 2 and may interface with one of more BS 6 and one of
more PSCN 8.
[0044] The Authentication, Authorization and Accounting (AAA) 13
Server provides Internet Protocol authentication, authorization and
accounting functions related to packet data traffic. The Home Agent
(HA) 14 provides authentication of MS 2 IP registrations, redirects
packet data to and from the Foreign Agent (FA) 15 component of the
PDSN 8, and receives provisioning information for users from the
AAA 13. The HA 14 may also establish, maintain, and terminate
secure communications to the PDSN 12 and assign a dynamic IP
address. The PDSN 12 communicates with the AAA 13, HA 14 and the
Internet 16 via an Internal IP Network.
[0045] There are several types of multiple access schemes,
specifically Frequency Division Multiple Access (FDMA), Time
Division Multiple Access (TDMA) and Code Division Multiple Access
(CDMA). In FDMA, user communications are separated by frequency,
for example, by using 30 KHz channels. In TDMA, user communications
are separated by frequency and time, for example, by using 30 KHz
channels with 6 timeslots. In CDMA, user communications are
separated by digital code.
[0046] In CDMA, All users on the same spectrum, for example, 1.25
MHz. Each user has a unique digital code identifier and the digital
codes separate users to prevent interference.
[0047] A CDMA signal uses many chips to convey a single bit of
information. Each user has a unique chip pattern, which is
essentially a code channel. In order to recover a bit, a large
number of chips are integrated according to a user's known chip
pattern. Other user's code patterns appear random and are
integrated in a self-canceling manner and, therefore, do not
disturb the bit decoding decisions made according to the user's
proper code pattern.
[0048] Input data is combined with a fast spreading sequence and
transmitted as a spread data stream. A receiver uses the same
spreading sequence to extract the original data. FIG. 2A
illustrates the spreading and de-spreading process. As illustrated
in FIG. 2B, multiple spreading sequences may be combined to create
unique, robust channels.
[0049] A Walsh code is one type of spreading sequence. Each Walsh
code is 64 chips long and is precisely orthogonal to all other
Walsh codes. The codes are simple to generate and small enough to
be stored in read only memory (ROM).
[0050] A short PN code is another type of spreading sequence. A
short PN code consists of two PN sequences (I and Q), each of which
is 32,768 chips long and is generated in similar, but differently
tapped 15-bit shift registers. The two sequences scramble the
information on the I and Q phase channels.
[0051] A long PN code is another type of spreading sequence. A long
PN code is generated in a 42-bit register and is more than 40 days
long, or about 4.times.10.sup.13 chips long. Due to its length, a
long PN code cannot be stored in ROM in a terminal and, therefore,
is generated chip-by-chip.
[0052] Each MS 2 codes its signal with the PN long code and a
unique offset, or public long code mask, computed using the long PN
code ESN of 32-bits and 10 bits set by the system. The public long
code mask produces a unique shift. Private long code masks may be
used to enhance privacy. When integrated over as short a period as
64 chips, MS 2 with different long PN code offsets will appear
practically orthogonal.
[0053] CDMA communication uses forward channels and reverse
channels. A forward channel is utilized for signals from a BTS 3 to
a MS 2 and a reverse channel is utilized for signals from a MS to a
BTS.
[0054] A forward channel uses its specific assigned Walsh code and
a specific PN offset for a sector, with one user able to have
multiple channel types at the same time. A forward channel is
identified by its CDMA RF carrier frequency, the unique short code
PN offset of the sector and the unique Walsh code of the user. CDMA
forward channels include a pilot channel, sync channel, paging
channels and traffic channels.
[0055] The pilot channel is a "structural beacon" which does not
contain a character stream, but rather is a timing sequence used
for system acquisition and as a measurement device during handoffs.
A pilot channel uses Walsh code 0.
[0056] The sync channel carries a data stream of system
identification and parameter information used by MS 2 during system
acquisition. A sync channel uses Walsh code 32.
[0057] There may be from one to seven paging channels according to
capacity requirements. Paging channels carry pages, system
parameter information and call setup orders. Paging channels use
Walsh codes 1-7.
[0058] The traffic channels are assigned to individual users to
carry call traffic, Traffic channels use any remaining Walsh codes
subject to overall capacity as limited by noise.
[0059] A reverse channel is utilized for signals from a MS 2 to a
BTS 3 and uses a Walsh code and offset of the long PN sequence
specific to the MS, with one user able to transmit multiple types
of channels simultaneously. A reverse channel is identified by its
CDMA RF carrier frequency and the unique long code PN Offset of the
individual MS 2. Reverse channels include traffic channels and
access channels,
[0060] Individual users use traffic channels during actual calls to
transmit traffic to the BTS 3, A reverse traffic channel is
basically a user-specific public or private long code Mask and
there are as many reverse traffic channels as there are CDMA
terminals,
[0061] An MS 2 not yet involved in a call uses access channels to
transmit registration requests, call setup requests, page
responses, order responses and other signaling information. An
access channel is basically a public long code offset unique to a
BTS 3 sector. Access channels are paired with paging channels, with
each paging channel having up to 32 access channels.
[0062] CDMA communication provides many advantages. Some of the
advantages are variable rate vocoding and multiplexing, power
control, use of RAKE receivers and soft handoff.
[0063] CDMA allows the use of variable rate vocoders to compress
speech, reduce bit rate and greatly increase capacity. Variable
rate vocoding provides full bit rate during speech, low data rates
during speech pauses, increased capacity and natural sound,
Multiplexing allows voice, signaling and user secondary data to be
mixed in CDMA frames.
[0064] By utilizing forward power control, the BTS 3 continually
reduces the strength of each user's forward baseband chip stream.
When a particular MS 2 experiences errors on the forward link, more
energy is requested and a quick boost of energy is supplied after
which the energy is again reduced,
[0065] Using a RAKE receiver allows a MS 2 to use the combined
outputs of the three traffic correlators, or "RAKE fingers," every
frame. Each RAKE finger can independently recover a particular PN
Offset and Walsh code. The fingers may be targeted on delayed
multipath reflections of different BTS's 3, with a searcher
continuously checking pilot signals.
[0066] The MS 2 drives soft handoff. The MS 2 continuously checks
available pilot signals and reports to the BTS 3 regarding the
pilot signals it currently sees. The BTS 3 assigns up to a maximum
of six sectors and the MS 2 assigns its fingers accordingly. A1
messages are sent by dim-and-burst without muting. Each end of the
communication link chooses the best configuration on a
frame-by-frame basis, with handoff transparent to users.
[0067] A cdma2000 system is a third-generation (3G) wideband;
spread spectrum radio interface system that uses the enhanced
service potential of CDMA technology to facilitate data
capabilities, such as Internet and intranet access, multimedia
applications, high-speed business transactions, and telemetry. The
focus of cdma2000 as is that of other third-generation systems, is
on network economy and radio transmission design to overcome the
limitations of a finite amount of radio spectrum availability.
[0068] FIG. 3 illustrates a data link protocol architecture layer
20 for a cdma2000 wireless network. The data link protocol
architecture layer 20 includes an Upper Layer 60, a Link Layer 30
and a Physical layer 21.
[0069] The Upper layer 60 includes three sublayers; a Data Services
sublayer 61; a Voice Services sublayer 62 and a Signaling Services
sublayer 63. Data services 61 are services that deliver any form of
data on behalf of a mobile end user and include packet data
applications such as IP service, circuit data applications such as
asynchronous fax and B-ISDN emulation services, and SMS. Voice
services 62 include PSTN access, mobile-to-mobile voice services,
and Internet telephony. Signaling 63 controls all aspects of mobile
operation.
[0070] The Signaling Services sublayer 63 processes all messages
exchanged between the MS 2 and BS 6. These messages control such
functions as call setup and teardown, handoffs, feature activation,
system configuration, registration and authentication.
[0071] The Link Layer 30 is subdivided into the Link Access Control
(LAC) sublayer 32 and the Medium Access Control (MAC) sublayer 31.
The Link Layer 30 provides protocol support and control mechanisms
for data transport services and performs the functions necessary to
map the data transport needs of the Upper layer 60 into specific
capabilities and characteristics of the Physical Layer 21. The Link
Layer 30 may be viewed as an interface between the Upper Layer 60
and the Physical Layer 20.
[0072] The separation of MAC 31 and LAC 32 sublayers is motivated
by the need to support a wide range of Upper Layer 60 services and
the requirement to provide for high efficiency and low latency data
services over a wide performance range, specifically from 1.2 Kbps
to greater than 2 Mbps. Other motivators are the need for
supporting high Quality of Service (QoS) delivery of circuit and
packet data services, such as limitations on acceptable delays
and/or data BER (bit error rate), and the growing demand for
advanced multimedia services each service having a different QoS
requirements.
[0073] The LAC sublayer 32 is required to provide a reliable,
in-sequence delivery transmission control function over a
point-to-point radio transmission link 42. The LAC sublayer 32
manages point-to point communication channels between upper layer
60 entities and provides framework to support a wide range of
different end-to-end reliable Link Layer 30 protocols.
[0074] The Link Access Control (LAC) sublayer 32 provides correct
delivery of signaling messages. Functions include assured delivery
where acknowledgement is required, unassured delivery where no
acknowledgement is required, duplicate message detection, address
control to deliver a message to an individual MS 2, segmentation of
messages into suitable sized fragments for transfer over the
physical medium, reassembly and validation of received messages and
global challenge authentication.
[0075] The MAC sublayer 31 facilitates complex multimedia,
multi-services capabilities of 3G wireless systems with QoS
management capabilities for each active service. The MAC sublayer
31 provides procedures for controlling the access of packet data
and circuit data services to the Physical Layer 21, including the
contention control between multiple services from a single user, as
well as between competing users in the wireless system. The MAC
sublayer 31 also performs mapping between logical channels and
physical channels, multiplexes data from multiple sources onto
single physical channels and provides for reasonably reliable
transmission over the Radio Link Layer using a Radio Link Protocol
(RLP) 33 for a best-effort level of reliability, Signaling Radio
Burst Protocol (SRBP) 35 is an entity that provides connectionless
protocol for signaling messages. Multiplexing and QoS Control 34 is
responsible for enforcement of negotiated QoS levels by mediating
conflicting requests from competing services and the appropriate
prioritization of access requests.
[0076] The Physical Layer 20 is responsible for coding and
modulation of data transmitted over the air. The Physical Layer 20
conditions digital data from the higher layers so that the data may
be transmitted over a mobile radio channel reliably
[0077] The Physical Layer 20 maps user data and signaling, which
the MAC sublayer 31 delivers over multiple transport channels, into
a physical channels and transmits the information over the radio
interface. In the transmit direction, the functions performed by
the Physical Layer 20 include channel coding, interleaving,
scrambling, spreading and modulation. In the receive direction, the
functions are reversed in order to recover the transmitted data at
the receiver.
[0078] FIG. 4 illustrates an overview of call processing.
Processing a call includes pilot and sync channel processing,
paging channel processing, access channel processing and traffic
channel processing.
[0079] Pilot and sync channel processing refers to the MS 2
processing the pilot and sync channels to acquire and synchronize
with the CDMA system in the MS 2 Initialization State. Paging
channel processing refers to the MS 2 monitoring the paging channel
or the forward common control channel (F-CCCH) to receive overhead
and mobile-directed messages from the BS 6 in the Idle State.
Access channel processing refers to the MS 2 sending messages to
the BS 6 on the access channel or the Enhanced access channel in
the System Access State, with the BS 6 always listening to these
channels and responding to the MS on either a paging channel or the
F-CCCH. Traffic channel processing refers to the BS 6 and MS 2
communicating using dedicated forward and reverse traffic channels
in the MS 2 Control on Traffic Channel State, with the dedicated
forward and reverse traffic channels carrying user information,
such as voice and data.
[0080] FIG. 5 illustrates the initialization state of a MS 2. The
Initialization state includes a System Determination Substate,
Pilot Channel Acquisition, Sync Channel Acquisition, a Timing
Change Substate and a Mobile Station Idle State.
[0081] System Determination is a process by which the MS 2 decides
from which system to obtain service. The process could include
decisions such as analog versus digital, cellular versus PCS, and A
carrier versus B carrier. A custom selection process may control
System Determination. A service provider using a redirection
process may also control System determination. After the MS 2
selects a system, it must determine on which channel within that
system to search for service. Generally the MS 2 uses a prioritized
channel list to select the channel.
[0082] Pilot Channel Processing is a process whereby the MS 2 first
gains information regarding system timing by searching for usable
pilot signals. Pilot channels contain no information, but the MS 2
can align its own timing by correlating with the pilot channel.
Once this correlation is completed, the MS 2 is synchronized with
the sync channel and can read a sync channel message to further
refine its timing. The MS 2 is permitted to search up to 15 seconds
on a single pilot channel before it declares failure and returns to
System Determination to select either another channel or another
system. The searching procedure is not standardized, with the time
to acquire the system depending on implementation.
[0083] In cdma2000, there may be many pilot channels, such as OTD
pilot, STS pilot and Auxiliary pilot, on a single channel. During
System Acquisition, the MS 2 will not find any of these pilot
channels because they are use different Walsh codes and the MS is
only searching for Walsh 0.
[0084] The sync channel message is continuously transmitted on the
sync channel and provides the MS 2 with the information to refine
timing and read a paging channel. The mobile receives information
from the BS 6 in the sync channel message that allows it to
determine whether or not it will be able to communicate with that
BS.
[0085] In the Idle State, the MS 2 receives one of the paging
channels and processes the messages on that channel. Overhead or
configuration messages are compared to stored sequence numbers to
ensure the MS 2 has the most current parameters. Messages to the MS
2 are checked to determine the intended subscriber.
[0086] The BS 6 may support multiple paging channels and/or
multiple CDMA channels (frequencies). The MS 2 uses a hash function
based on its IMSI to determine which channel and frequency to
monitor in the Idle State. The BS 6 uses the same hash function to
determine which channel and frequency to use when paging the MS
2.
[0087] Using a Slot Cycle Index (SCI) on the paging channel and on
F-CCCH supports slotted paging. The main purpose of slotted paging
is to conserve battery power in MS 2. Both the MS 2 and BS 6 agree
in which slots the MS will be paged. The MS 2 can power down some
of its processing circuitry during unassigned slots, Either the
General Page message or the Universal Page message may be used to
page the mobile on F-CCCH. A Quick paging channel that allows the
MS 2 to power up for a shorter period of time than is possible
using only slotted paging on F-PCH or F-CCCH is also supported.
[0088] FIG. 6 illustrates the System Access state. The first step
in the system access process is to update overhead information to
ensure that the MS 2 is using the correct access channel
parameters, such as initial power level and power step increments.
A MS 2 randomly selects an access channel and transmits without
coordination with the BS 6 or other MS. Such a random access
procedure can result in collisions. Several steps can be taken to
reduce the likelihood of collision, such as use of a slotted
structure, use of a multiple access channel, transmitting at random
start times and employing congestion control, for example, overload
classes.
[0089] The MS 2 may send either a request or a response message on
the access channel. A request is a message sent autonomously, such
as an Origination message. A response is a message sent in response
to a message received from the BS 6. For example, a Page Response
message is a response to a General Page message or a Universal
message.
[0090] An access attempt, which refers to the entire process of
sending one Layer 2 encapsulated PDU and receiving an
acknowledgment for the PDU, consists of one or more access
sub-attempts, as illustrated in FIG. 7. An access sub-attempt
includes of a collection of access probe sequences, as illustrated
in FIG. 8. Sequences within an access sub-attempt are separated by
a random backoff interval (RS) and a persistence delay (PD). PD
only applies to access channel request, not response. FIG. 9
illustrates a System Access state in which collisions are avoided
by using a slot offset of 0-511 slots.
[0091] The Multiplexing and QoS Control sublayer 34 has both a
transmitting function and a receiving function. The transmitting
function combines information from various sources, such as Data
Services 61, Signaling Services 63 or Voice Services 62, and forms
Physical layer SDUs and PDCHCF SDUs for transmission. The receiving
function separates the information contained in Physical Layer 21
and PDCHCF SDUs and directs the information to the correct entity,
such as Data Services 61, Upper Layer Signaling 63 or Voice
Services 62.
[0092] The Multiplexing and QoS Control sublayer 34 operates in
time synchronization with the Physical Layer 21. If the Physical
Layer 21 is transmitting with a non-zero frame offset, the
Multiplexing and QoS Control sublayer 34 delivers Physical Layer
SDUs for transmission by the Physical Layer at the appropriate
frame offset from system time.
[0093] The Multiplexing and QoS Control sublayer 34 delivers a
Physical Layer 21 SDU to the Physical Layer using a
physical-channel specific service interface set of primitives. The
Physical Layer 21 delivers a Physical Layer SDU to the Multiplexing
and QoS Control sublayer 34 using a physical channel specific
Receive Indication service interface operation.
[0094] The SRBP Sublayer 35 includes the sync channel, forward
common control channel, broadcast control channel, paging channel
and access channel procedures.
[0095] The LAC Sublayer 32 provides services to Layer 3 60. SDUs
are passed between Layer 3 60 and the LAC Sublayer 32. The LAC
Sublayer 32 provides the proper encapsulation of the SDUs into LAC
PDUs, which are subject to segmentation and reassembly and are
transferred as encapsulated PDU fragments to the MAC Sublayer
31.
[0096] Processing within the LAC Sublayer 32 is done sequentially,
with processing entities passing the partially formed LAC PDU to
each other in a well-established order. SDUs and PDUs are processed
and transferred along functional paths, without the need for the
upper layers to be aware of the radio characteristics of the
physical channels. However, the upper layers could be aware of the
characteristics of the physical channels and may direct Layer 2 30
to use certain physical channels for the transmission of certain
PDUs.
[0097] A 1xEV-DO system is optimized for packet data service and
characterized by a single 1.25 MHz carrier ("1x") for data only or
data Optimized ("DO"). Furthermore, there is a peak data rate of
2.4 Mbps or 3.072 Mbps on the forward Link and 153.6 Kbps or 1.8432
Mbps on the reverse Link. Moreover, a 1xEV-DO system provides
separated frequency bands and internetworking with a 1x System.
FIG. 10 illustrates a comparison of cdma2000 for a 1x system and a
1xEV-DO system.
[0098] In CDMA2000, there are concurrent services, whereby voice
and data are transmitted together at a maximum data rate of 614.4
kbps and 307.2 kbps in practice. An MS 2 communicates with the MSC
5 for voice calls and with the PDSN 12 for data calls. A cdma2000
system is characterized by a fixed rate with variable power with a
Walsh-code separated forward traffic channel.
[0099] In a 1xEV-DO system, the maximum data rate is 2.4 Mbps or
3.072 Mbps and there is no communication with the circuit-switched
core network 7. A 1xEV-DO system is characterized by fixed power
and a variable rate with a single forward channel that is time
division multiplexed.
[0100] FIG. 11 illustrates a 1xEV-DO system architecture. In a
1xEV-DO system, a frame consists of 16 slots, with 600 slots/sec,
and has a duration of 26.67 ms, or 32,768 chips. A single slot is
1.6667 ms long and has 2048 chips. A control/traffic channel has
1600 chips in a slot, a pilot channel has 192 chips in a slot and a
MAC channel has 256 chips in a slot. A 1xEV-DO system facilitates
simpler and faster channel estimation and time synchronization,
[0101] FIG. 12 illustrates a 1xEV-DO default protocol architecture.
FIG. 13 illustrates a 1xEV-DO non-default protocol
architecture.
[0102] Information related to a session in a 1xEV-DO system
includes a set of protocols used by an MS 2, or access terminal
(AT), and a BS 6, or access network (AN), over an airlink, a
Unicast Access Terminal Identifier (UATI), configuration of the
protocols used by the AT and AN over the airlink and an estimate of
the current AT location.
[0103] The Application Layer provides best effort, whereby the
message is sent once, and reliable delivery, whereby the message
can be retransmitted one or more times. The stream layer provides
the ability to multiplex up to 4 (default) or 255 (non-default)
application streams for one AT 2.
[0104] The Session Layer ensures the session is still valid and
manages closing of session, specifies procedures for the initial
UATI assignment, maintains AT addresses and negotiates/provisions
the protocols used during the session and the configuration
parameters for these protocols.
[0105] FIG. 14 illustrates the establishment of a 1xEV-DO session.
As illustrated in FIG. 14, establishing a session includes address
configuration, connection establishment, session configuration and
exchange keys.
[0106] Address configuration refers to an Address Management
protocol assigning a UATI and Subnet mask. Connection establishment
refers to Connection Layer Protocols setting up a radio link.
Session configuration refers to a Session Configuration Protocol
configuring all protocols. Exchange key refers a Key Exchange
protocol in the Security Layer setting up keys for
authentication.
[0107] A "session` refers to the logical communication link between
the AT 2 and the RNC, which remains open for hours, with a default
of 54 hours. A session lasts until the PPP session is active as
well. Session information is controlled and maintained by the RNC
in the AN 6.
[0108] When a connection is opened, the AT 2 can be assigned the
forward traffic channel and is assigned a reverse traffic channel
and reverse power control channel. Multiple connections may occur
during single session.
[0109] The Connection Layer manages initial acquisition of the
network and communications. Furthermore, the Connection Layer
maintains an approximate AT 2 location and manages a radio link
between the AT 2 and the AN 6. Moreover, the Connection Layer
performs supervision, prioritizes and encapsulates transmitted data
received from the Session Layer, forwards the prioritized data to
the Security Layer and decapsulates data received from the Security
Layer and forwards it to the Session Layer.
[0110] FIG. 15 illustrates Connection Layer Protocols.
[0111] In the Initialization State, the AT 2 acquires the AN 6 and
activates the initialization State Protocol, In the Idle State, a
closed connection is initiated and the Idle State Protocol is
activated. In the Connected State, an open connection is initiated
and the Connected State Protocol is activated.
[0112] A closed connection refers to a state where the AT 2 is not
assigned any dedicated air-link resources and communications
between the AT and AN 6 are conducted over the access channel and
the control channel. An open connection refers to a state where the
AT 2 can be assigned the forward traffic channel, is assigned a
reverse power control channel and a reverse traffic channel and
communication between the AT 2 and AN 6 is conducted over these
assigned channels as well as over the control channel.
[0113] The Initialization State Protocol performs actions
associated with acquiring an AN 6. The Idle State Protocol performs
actions associated with an AT 2 that has acquired an AN 6, but does
not have an open connection, such as keeping track of the AT
location using a Route Update Protocol. The Connected State
Protocol performs actions associated with an AT 2 that has an open
connection, such as managing the radio link between the AT and AN 6
and managing the procedures leading to a closed connection. The
Route Update Protocol performs actions associated with keeping
track of the AT 2 location and maintaining the radio link between
the AT and AN 6. The Overhead Message Protocol broadcasts essential
parameters, such as QuickConfig, SectorParameters and
AccessParameters message, over the control channel, The Packet
Consolidation Protocol consolidates and prioritizes packets for
transmission as a function of their assigned priority and the
target channel as well as providing packet de-multiplexing on the
receiver.
[0114] The Security Layer includes a key exchange function,
authentication function and encryption function. The key exchange
function provides the procedures followed by the AN 2 and AT 6 for
authenticating traffic. The authentication function provides the
procedures followed by the AN 2 and AT 6 to exchange security keys
for authentication and encryption. The encryption function provides
the procedures followed by the AN 2 and AT 6 for encrypting
traffic.
[0115] The 1xEV-DO forward Link is characterized in that no power
control and no soft handoff is supported. The AN 6 transmits at
constant power and the AT 2 requests variable rates on the forward
Link. Because different users may transmit at different times in
TDM, it is difficult to implement diversity transmission from
different BS's 6 that are intended for a single user.
[0116] In the MAC Layer, two types of messages originated from
higher layers are transported across the physical layer,
specifically a User data message and a signaling message. Two
protocols are used to process the two types of messages,
specifically a forward traffic channel MAC Protocol for the User
data message and a control channel MAC Protocol, for the signaling
message.
[0117] The Physical Layer is characterized by a spreading rate of
1.2288 Mcps, a frame consisting of 16 slots and 26.67 ms, with a
slot of 1.67 ms and 2048 chips. The forward Link channel includes a
pilot channel, a forward traffic channel or control channel and a
MAC channel,
[0118] The pilot channel is similar to the to the cdma2000 pilot
channel in that it comprises all "0" information bits and
Walsh-spreading with WO with 192 chips for a slot.
[0119] The forward traffic channel is characterized by a data rate
that varies from 38.4 kbps to 2.4576 Mbps or from 4.8 kbps to 3.072
Mbps. Physical Layer packets can be transmitted in 1 to 16 slots
and the transmit slots use 4-slot interlacing when more than one
slot is allocated. If ACK is received on the reverse Link ACK
channel before all of the allocated slots have been transmitted the
remaining slots shall not be transmitted.
[0120] The control channel is similar to the sync channel and
paging channel in cdma2000. The control channel is characterized by
a period of 256 slots or 427.52 ms, a Physical Layer packet length
of 1024 bits or 128, 256, 512 and 1024 bits and a data rate of 38.4
kbps or 76.8 kbps or 19.2 kbps, 38.4 kbps or 76.8 kbps.
[0121] The 1xEV-DO reverse link is characterized in that the AN 6
can power control the reverse Link by using reverse power control
and more than one AN can receive the AT's 2 transmission via soft
handoff. Furthermore, there is no TDM on the reverse Link, which is
channelized by Walsh code using a long PN code.
[0122] An access channel is used by the AT 2 to initiate
communication with the AN 6 or to respond to an AT directed
message. Access channels include a pilot channel and a data
channel.
[0123] An AT 2 sends a series of access probes on the access
channel until a response is received from the AN 6 or a timer
expires. An access probe includes a preamble and one or more access
channel Physical Layer packets. The basic data rate of the access
channel is 9.6 kbps, with higher data rates of 19.2 kbps and 38.4
kbps available.
[0124] When more that one AT 2 is paged using the same Control
channel packet, Access Probes may be transmitted at the same time
and packet collisions are possible. The problem can be more serious
when the ATs 2 are co-located, are in a group call or have similar
propagation delays.
[0125] One reason for the potential of collision is the
inefficiency of the current persistence test in conventional
methods. Because an AT 2 may require a short connection setup time,
a paged AT may transmit access probes at the same time as another
paged AT when a persistence test is utilized.
[0126] Conventional methods that use a persistence test are not
sufficient since each AT 2 that requires a short connection setup
times and/or is part of a group call may have the same persistence
value, typically set to 0. If AT's 2 are co-located, such as In a
group call, the Access Probes arrive at the An 6 at the same time,
thereby resulting in access collisions and increased connection
setup time.
[0127] Therefore, there is a need for a more efficient approach for
access probe transmission from co-located mobile terminals
requiring short connection times. The present invention addresses
this and other needs such as interference cancellation.
[0128] Multi-user detection strategy relates to a method for
mitigating multiple access interference (MAI) effects and solving
near-far problem with exploiting interference structure. Researches
are being conducted to reduce computation complexity and prior
knowledge using blind multi-user detection and subspace-based
signature waveform estimation.
[0129] Blind multi-user detectors can be used to achieve good
performance with only the knowledge of desired users' timing and
signature waveform.
[0130] There are two popular approaches for designing blind
multi-user detectors. One of the approaches is to use the
conventional multi-user signal model in which received signals and
multi-user receivers are taken as linear combinations of actual
spreading sequences and noise and statistical signal estimation
techniques for blind multi-user detection (e.g., blind multi-user
receiver design using Wiener filters or Kalman filters
techniques).
[0131] The other approach is based on parametric signal modeling
and signal spectrum estimation, where received signals and
multi-user receivers are taken as a linear combination of desired
users' spreading sequences and the signal/noise subspace bases.
Many subspace-based schemes are examples of this approach, which
essentially is a method for blindly reconstructing existing
conventional multi-user detectors using subspace concept.
[0132] Although both the conventional multi-user signal model and
subspace-based signal model provide natural and straightforward
representations of received signals and multi-user detectors, the
blind detectors based on these two models are difficult to be
implemented in practice since these two models are not known
beforehand and difficult to accurately estimate them. Moreover,
computations associated with these two models are highly complex,
especially for practical applications.
[0133] To address the near-far problem with minimum prior knowledge
and computation complexity, an alternative blind multi-user
framework based on a blind multi-user signal model can be used.
Based on the alternative blind multi-user signal model and
detection framework, blind multi-user detectors can be developed
using best linear unbiased and minimum means squared error
estimation (MMSE) criteria in addition to a least squares based
scheme. Here, the associated algorithms are simple and direct by
only using the signatures and timing of desired users, Further,
there is no statistical signal estimation or subspace separation
procedure employed by many other blind detectors.
[0134] Compared to the existing blind detection schemes which use
all of the received signals, the alternative blind multi-user
signal model requires a minimum number of previously received
signals. As such, the computation complexity and detection delay
can be noticeably reduced. A recursively adaptive implementation is
provided to further lessen the required computation complexity.
[0135] Hereafter, details of performance related to the near-far
problem as well as the trade-off between the complexity and
performances will be discussed.
[0136] In the discussion below, a forward link (FL) transmission in
a single-cell DS/CDMA system is assumed. There are K active users
over the multi-path channel with P strong paths. Here, the strong
paths are those to be explicitly combined by RAKE receiver.
Moreover, the channel is an additive white Gaussian noise (AWGN)
channel. The baseband representation of the received signal due to
user k is given by the following equation. r k .function. ( t ) = p
= 1 P .times. .alpha. pk .times. A k .function. [ n ] .times. b k
.function. [ n ] .times. c k .function. ( t - n .times. .times. T -
.tau. p ) [ Equation .times. .times. 1 ] ##EQU1##
[0137] In Equation 1, .alpha..sub.pk is the pth path loss of user
k's signal, b.sub.k[n] is the nth bit sent by user k. Here,
assumption can be made that {b.sub.k[n]} are independent and
identically distributed random variables with E{b.sub.k[i]}=0 and
E{b.sub.k[i]|.sup.2}=1. The parameters c.sub.k(t) denote the
normalized spreading signal waveform of user k during the interval
[0, T], 0.ltoreq..tau..sub.1.ltoreq..tau..sub.2 . . .
.ltoreq..tau..sub.p denotes P different transmission delays from
the base station to user k, and A.sub.k[n] is the received signal
amplitude for user k at time t=n, which depends on the possible
channel statistics.
[0138] The total baseband signal received by user k is represented
by the following equation, r _ .function. ( t ) = k = 1 K .times. r
k .function. ( t ) [ Equation .times. .times. 2 ] ##EQU2##
[0139] According to Equation 2, the received signal {overscore
(r)}(t) is passed through the corresponding chip matched filter
(CMF), .phi.(t), and RAKE combiner, The combined output r(t) is
represented by Equation 3.
r(t)=A.sub.kb.sub.kc.sub.k(t-nT-.tau..sub.1){circle around
(.times.)}.phi.(t-.tau..sub.1)+m.sub.ISI(t)+m.sub.MAI(t)+n(t)
[Equation 3]
[0140] In Equation 3, m ISI .function. ( t ) = p .noteq. q P
.times. .beta. qk .times. .alpha. p .times. .times. k .times. A k
.times. b k .times. c k .function. ( t - n .times. .times. T +
.tau. q .times. .times. 1 - .tau. 1 ) .PHI. .function. ( t - .tau.
1 ) ##EQU3## is the inter-symbol interference (ISI) to user k.
Furthermore, the MAI to user k is represented as m MAI .function. (
t ) = .times. i .noteq. k K .times. A i .times. b i .times. c i
.function. ( t - n .times. .times. T - .tau. 1 ) .PHI. .function. (
t - .tau. 1 ) + .times. i .noteq. k K .times. p .noteq. q P .times.
.beta. q .times. .times. k .times. .alpha. p .times. .times. t
.times. A t .times. b t .times. c t .function. ( t - n .times.
.times. T + .tau. q .times. .times. 1 - .tau. p ) .PHI. .function.
( t - .tau. 1 ) ##EQU4## Here, .beta..sub.qk is the weight of the
qth RAKE finger with q = 1 P .times. .beta. q .times. .times. k
.times. .alpha. q .times. .times. k = 1 ##EQU5## and
.tau..sub.q1=.tau..sub.q-.tau..sub.1 is the propagation delay
difference between the 1.sup.st path and pth path. In addition,
{circle around (.times.)} denotes the convolutional product while
n(t) is AWGN with variance .sigma..sup.2. The user k's RAKE output
can be sampled at f.sub.s=1/T.sub.s. This can be expressed in a
straightforward manner as shown in Equation 4. r = .times. [ r
.function. ( n .times. .times. T + T s + .tau. 1 ) .times. .times.
.times. .times. r .function. ( n .times. .times. T + L .times.
.times. T s + .tau. 1 ) ] T = .times. k = 1 K .times. A k .times. B
k .times. s k + n = .times. S .times. .times. A .times. .times. b +
n [ Equation .times. .times. 4 ] ##EQU6##
[0141] In Equation 4, S=[s.sub.1 s.sub.2 . . . s.sub.K] is the
received signal signature matrix combined with both, and
L=T/T.sub.s is the number of sample per symbol, which usually is
not less than the spreading gain L.sub.c.
[0142] Because of m.sup.MAI(t) existing in the received signal
r(t), the performance of conventional matched filter receiver
suffers from the so-called near-far problem. Multi-user detection
is the receiver technique for solving this problem, and most
multi-user detectors are developed using the conventional system
model like the model of Equation 4. One of the difficulties is
developing blind multi-user detectors using Equation 4 is that S is
difficult to acquire in advance. With that, it takes much effort to
acquire S at a later time. Similar situation can arise in
developing blind detectors using parametric subspace signal
model.
Blind Multi-User Detection Framework
[0143] Instead of using the conventional signal model or the
parametric subspace signal model, a blind multi-user signal model
of FIG. 16 is introduced. FIG. 16 is an exemplary diagram of a
receiver design. In FIG. 16, a blind but "faked" spreading matrix S
is used. The blind but "faked" spreading matrix S is composed by
only desired users' spreading sequences and previously received
signals. Further, the blind but "faked" spreading matrix S is not
an original signal but functions as the original signal bases for
representing received signals and multi-user detectors.
[0144] With respect to construction of S, a grouped blind detector
or a single blind detector can be constructed.
[0145] In detail, without loss of generality, only the bits sent
for first G users are considered. Moreover, the blind spreading
sequence matrix S, in L.times.M format, can be defined by
S=[s.sub.1 . . . s.sub.G r.sub.1 r.sub.2 . . . r.sub.M-G], where
s.sub.g with g=1, 2, . . . , G, denote the group of G spreading
waveforms which are already known to user 1, Further, r.sub.m with
m=1, 2, . . . , M-G, are M-G previously received independent signal
vectors. Here, the first M-G received signal r.sub.m may be
obtained from some receiver initialization procedure. After that,
the may be some possible adaptive procedure for updating S.
Compared to existing blind detectors, this number, M-G, of required
previously received signals are very small.
[0146] In addition, K.ltoreq.M.ltoreq.L, where M=K is the minimum
number for blind detector to unambiguously distinguish different
interfering signals and M.ltoreq.L is the constraint for
guaranteeing the uniqueness of designed blind multi-user receiver.
The relationship between the proposed blind spreading matrix S and
the original spreading matrix S can be represented according to the
following equation. S=SB+N [Equation 5]
[0147] In Equation 5, the first G columns of S and S are same. The
following equation represents K.times.M data matrix associated with
S. B = [ I D _ 0 D .about. ] = [ E .times. D _ D .about. ] = [ G 0
D .about. ] [ Equation .times. .times. 6 ] ##EQU7##
[0148] E=[I 0].sup.T is a K.times.G matrix, G=[I {tilde over (D)}]
is the G.times.M matrix. Here, {tilde over (D)} is the previously
detected and known matrix for desired users. Moreover, rank {{tilde
over (D)}}=K-G where {tilde over (D)} is unknown matrix for unknown
K-G users while {B}.ltoreq.K. By combining Equations 4 and 5, the
received signal vector r in Equation 5 can be expressed as the
linear combination of the columns in S instead of S. The received
signal vector r can be written as shown by Equation 7. r=Sf+n
[Equation 7]
[0149] In Equation 7, M.times.1 vector f is termed the detection
vector defined by f=B.sup.+{overscore (b)} where [].sup.+ denotes
the general inverse operator and {overscore (b)}=Ab. Furthermore, n
represents the new L.times.1 AWGN vector defined by the following
equation. n=n-NB.sup.+n [Equation 8]
[0150] Referring to Equation 10, this equation can be taken as a
modified linear prediction model and the multi-user detection
problem may be taken as a modified linear prediction problem in
n=0. On the other hand, if f can be estimated, the amplitudes
A.sub.g and bits b.sub.g for the first G users can be estimated and
detected with Equation 7 by {circumflex over
(b)}.sub.1=[{circumflex over (b)}.sub.1 {circumflex over (b)}.sub.2
. . . {circumflex over (b)}.sub.G].sup.T=sgn{Gf} and
a.sub.1=[a.sub.1 a.sub.2 . . . a.sub.G].sup.T=sgn|Gf|.
[0151] There are various schemes by which the values of f and
b.sub.1 can be estimated and detected. Among many available, a
least squares (LS) detection scheme can be used. Here, an
assumption is made that the measurements of S are free of error,
All errors are confined to the receiver vector r. Hence, the
detection vector can be estimated with solving the following
equation. f LS = .times. arg .times. .times. min x .times. r - Sx 2
= .times. S + .times. r [ Equation .times. .times. 9 ] ##EQU8##
[0152] Further, the bit vector for the first G users can be
detected by the following equation. {circumflex over
(b)}.sub.1=sign{GS.sup.+r} [Equation 10]
[0153] In addition to the LS detection scheme, a total least
squares (TLS) detection scheme can be used. The previous LS
estimation assumes S to be error-free. This assumption is not
entirely accurate with S because of N. As such, Equation 9 can be
transformed into the TLS problem as shown in Equation 11. [ S TLS f
TLS ] = arg .times. .times. min S , x .times. [ S r ] - [ S _ S _ x
] 2 [ Equation .times. .times. 11 ] ##EQU9##
[0154] Let S=U'.SIGMA.'V'.sup.T and [S r]=U.SIGMA.V.sup.T be the
SVD of S and [S r], respectively. If
.sigma.'.sub.K>.sigma..sub.K+1, the TLS estimation of f can be
expressed according to the following equation.
f.sub.TLS=(S.sup.TS-.sigma..sub.K+1.sup.2I).sup.-1S.sup.Tr
[0155] Further, the bit vector for the first G users can be
detected by the following equation. {circumflex over
(b)}.sub.1=sign{G(S.sup.TS-.sigma..sub.K+1.sup.2I).sup.-1S.sup.Tr}
[Equation 13]
[0156] Alternatively, a mixed LS/TLS detection scheme can also be
implemented. More specifically, although there exists a noise or
error matrix N in S, its first G columns are exactly known to be
noise-free or error-free. Hence, to maximize the estimation
accuracy of the detection vector f, it is natural to require the
corresponding columns of S to be unperturbed since they are known
exactly. Equations 9 and 11 can be transformed into the following
MLS equations. [ S MLS f MLS ] = arg .times. .times. min S .about.
, y .times. [ S .about. r ] - [ Z [ s t .times. S ] .times. x ] 2 [
Equation .times. .times. 14 ] ##EQU10##
[0157] Thereafter, Householder transformation Q can be performed on
the matrix [S r] which results in the following equation. Q T
.function. [ s 1 .times. .times. .times. .times. s G .times. S
.about. .times. .times. r ] = [ R 11 R 12 r 1 .times. .times. r 0 R
22 r 2 .times. .times. r ] [ Equation .times. .times. 15 ]
##EQU11##
[0158] In Equation 15, R.sub.11 is a G.times.G upper triangle
matrix, r.sub.1r is a G.times.1 vector, and r.sub.2r is a
(L-G).times.1 vector. The smallest singular value of R.sub.22 can
be denoted as .sigma.', and the smallest singular value of
[R.sub.22 r.sub.2r] can be .sigma.. If .sigma.'>.sigma., then
the MLS solution uniquely exists and can be expressed by the
following equation. f MLS = ( S T .times. S - .sigma. 2 .function.
[ 0 0 0 I M - G ] ) - 1 .times. S T .times. r [ Equation .times.
.times. 16 ] ##EQU12##
[0159] Further, the bit vector for the first G users can be
detected by the following equation. b ^ 1 = sign .times. { G
.function. ( S T .times. S - .sigma. 2 .function. [ 0 0 0 I M - G ]
) - 1 .times. S T .times. r } [ Equation .times. .times. 17 ]
##EQU13##
[0160] As another alternative scheme, a best linear unbiased
detection (BLU) scheme can be used. In the BLU scheme, the linear
structure f.sub.BLU=W.sub.BLU.sup.Tr for this so-called best linear
unbiased estimator (BLUE), which is equal to the optimal minimum
variance unbiased estimator (MVUE) in linear signal models if the
data are truly Gaussian. Here, matrix W.sub.BLU is designed such
that S must be deterministic, {tilde over (n )} must be zero mean
with positive definite known covariance matrix f.sub.BLU, f.sub.BLU
is an unbiased estimator of f, and the error variance for each of
the M parameters is minimized as shown in the following equation. W
BLU = min .times. W f .times. var .times. { W r T .times. r } [
Equation .times. .times. 18 ] ##EQU14##
[0161] The resulting best linear unbiased estimator is represented
by the following equation.
f.sub.BLU=(S.sup.TC.sub.n.sup.-1S).sup.-1S.sup.TC.sub.n.sup.-1r tm
[Equation 19]
[0162] The covariance matrix of f.sub.BLU is given by the following
equation. C.sub.f.sub.BLU=(S.sup.TC.sub.n.sup.-1S).sup.-1 [Equation
20]
[0163] Although the PDF of B may be determined, the PDF of B.sup.+
is largely unknown, However, with Girko's Law, when
.alpha.=(K-G)/(M-G) is fixed, K, M.fwdarw..infin., the diagonal
element of 1 M - G .function. [ D ~ + .times. b ~ .times. b ~ T
.times. D ~ + T ] n ~ - 1 = 1 - .alpha. ##EQU15## may be
approximated by Equation 21. lim .times. 1 M - G .function. [ D ~ +
.times. b ~ .times. b ~ T .times. D ~ + T ] n ~ - 1 = 1 - .alpha. [
Equation .times. .times. 21 ] ##EQU16##
[0164] Hence, the covariance matrix of f, C.sub.f, can be decided
by the following equation. C f = [ 2 .times. M - K - G M - K
.times. A t 2 0 T 0 1 M - K .times. I ] [ Equation .times. .times.
22 ] ##EQU17##
[0165] Referring to Equation 22, A .times. .times. 1 = diag .times.
{ A ^ 1 A ^ 2 A ^ G } , C n ~ = .sigma. 2 .times. 2 .times. M - K -
G M - K .times. I , ##EQU18## and the bit vector for the first G
users can be detected by {circumflex over (b)}.sub.1=sign
{G(S.sup.TS).sup.-1S.sup.Tr}.
[0166] Lastly, another blind multi-user detector scheme can be a
minimum mean squared error (MMSE) detection scheme. Under this
scheme, give measurements r, the MMSE estimator of f (i.e.,
f.sub.MMSE=f(r)), minimizes the MSE
J.sub.MSE={.parallel.f-{circumflex over (f)}.parallel..sub.2
.sup.2}. When f and r are jointly Gaussian, the linear estimator
W.sub.MMSE that minimizes the MSE J.sub.MSE can be represented
according to the following equation.
f.sub.MMSE=(C.sub.r.sup.-1+S.sup.TC.sub.n.sup.-1S).sup.-1S.sup.TC.sub.n.s-
up.-1r [Equation 23]
[0167] This equation is also termed Wiener filter, and the bit
vector for the first G users can be detected by {circumflex over
(b)}.sub.1=sign{G(C.sub.r.sup.-1+S.sup.TC.sub.n.sup.-1S).sup.-1S.sup.TC.s-
ub.n.sup.-1r}.
[0168] With respect to the discussion of the blind multi-user
detector scheme according to the embodiment of the present
invention, it is possible to integrate a multi-input multi-output
(MIMO) techniques as well as a multi-user detection (MUD)
schemes.
[0169] It will be apparent to those skilled in the art that various
modifications and variations can be made in the present invention
without departing from the spirit or scope of the inventions. Thus,
it is intended that the present invention covers the modifications
and variations of this invention provided they come within the
scope of the appended claims and their equivalents.
* * * * *