U.S. patent application number 11/888125 was filed with the patent office on 2007-12-27 for uwb communication receiver feedback loop.
This patent application is currently assigned to The Regents of the University of California. Invention is credited to Dave Benzel, Farid U. Dowla, Faranak Nekoogar, Erwin T. Rosenbury, Alex Spiridon.
Application Number | 20070297543 11/888125 |
Document ID | / |
Family ID | 33457553 |
Filed Date | 2007-12-27 |
United States Patent
Application |
20070297543 |
Kind Code |
A1 |
Spiridon; Alex ; et
al. |
December 27, 2007 |
UWB communication receiver feedback loop
Abstract
A novel technique and structure that maximizes the extraction of
information from reference pulses for UWB-TR receivers is
introduced. The scheme efficiently processes an incoming signal to
suppress different types of UWB as well as non-UWB interference
prior to signal detection. Such a method and system adds a feedback
loop mechanism to enhance the signal-to-noise ratio of reference
pulses in a conventional TR receiver. Moreover, sampling the second
order statistical function such as, for example, the
autocorrelation function (ACF) of the received signal and matching
it to the ACF samples of the original pulses for each transmitted
bit provides a more robust UWB communications method and system in
the presence of channel distortions.
Inventors: |
Spiridon; Alex; (Palo Alto,
CA) ; Benzel; Dave; (Livermore, CA) ; Dowla;
Farid U.; (Castro Valley, CA) ; Nekoogar;
Faranak; (San Ramon, CA) ; Rosenbury; Erwin T.;
(Castro Valley, CA) |
Correspondence
Address: |
Michael C. Staggs;Attorney for Applicants
Lawrence Livermore National Laboratory
P.O. Box 808, L-703
Livermore
CA
94551
US
|
Assignee: |
The Regents of the University of
California
|
Family ID: |
33457553 |
Appl. No.: |
11/888125 |
Filed: |
July 30, 2007 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10815450 |
Mar 31, 2004 |
|
|
|
11888125 |
Jul 30, 2007 |
|
|
|
60474150 |
May 28, 2003 |
|
|
|
Current U.S.
Class: |
375/345 ;
375/E7.067 |
Current CPC
Class: |
H04B 1/719 20130101 |
Class at
Publication: |
375/345 ;
375/E07.067 |
International
Class: |
H04L 27/08 20060101
H04L027/08 |
Goverment Interests
[0002] The United States Government has rights in this invention
pursuant to Contract No. W-7405-ENG-48 between the United States
Department of Energy and the University of California for the
operation of Lawrence Livermore National Laboratory.
Claims
1. An ultra-wideband (UWB) communication receiver, comprising: a
feedback amplifier having an input line and an output line, said
input line electrically coupled to a common input line, wherein
said feedback amplifier, said input line and said output line are
configured as a feedback loop so as to loop at least once, one or
more transmitted pulse pairs received on said common input line,
wherein said transmitted pulse pairs have a predetermined encoded
data and a predetermined symbol repetition rate; at least one delay
electrically coupled to said common input line and configured to
reproduce a predetermined lag interval D of said one or more
transmitted pulse pairs; a signal multiplier electrically coupled
to said at least one delay and said output line of said feedback
loop for multiplying delayed versions having said lag interval D of
said transmitted pulse pairs and one or more of said looped
transmitted pulse pairs; and an integrator configured to integrate
a product signal output by said signal multiplier, wherein an
integrator output is analyzed to determine an output signal that
indicates said predetermined encoded data.
2. The receiver of claim 1, wherein a gain of said feedback loop is
less than 1.
3. The receiver of claim 1, wherein said feedback loop averages out
interfering narrowband signals and white noise by adjusting a
feedback loop travel time delay to substantially match said
predetermined symbol repetition rate of said received pulse
pairs.
4. The receiver of claim 1, wherein said encoding includes a
relative polarity of a received transmitted data pulse with respect
to a received transmitted reference pulse.
Description
RELATED APPLICATION
[0001] This application is a continuation of prior application Ser.
No. 10/815,450 filed Mar. 31, 2004, entitled "UWB Communications
Receiver Feedback Loop", which claims the benefit of U.S.
Provisional Application No. 60/474,150, filed May 28, 2003, and
entitled, "Impulse Communications", both of which are incorporated
herein by this reference. Any disclaimer that may have occurred
during the prosecution of the above-referenced application Ser. No.
10/815,450 is hereby expressly rescinded.
BACKGROUND OF THE INVENTION
[0003] 1. Field of the Invention
[0004] The present invention relates to a communication method,
apparatus and system, and more particularly to an ultra wide-band
(UWB) communication apparatus, system and method employing a delay
loop to suppress various types of channel noise, such as, additive
white Gaussian noise (AWGN) or narrowband interference (NBI) on the
transmitted signal.
[0005] 2. Description of Related Art
[0006] Ultra-wideband (UWB) technology provides the potential for
delivering a large amount of data with low power spectral density
due to modulation of extremely narrow pulses. The short duration of
UWB pulses spreads their energy across a wide range of frequencies
from near DC to several GHz and enables UWB signals to share the
frequency spectrum with the coexisting narrowband and wideband
communication systems. Although UWB communications offer a
promising solution to an increasingly overcrowded frequency
spectrum, the overlay of UWB signals on coexisting narrowband
systems can impart interference, i.e., noise, for both UWB and
narrowband systems. Studies show that UWB systems are not
considered a serious threat to most narrowband systems due to their
low power spectral densities. However, strong interference from
narrowband systems on UWB devices can be detrimental to low powered
UWB waveforms and is not to be overlooked. Therefore, proper
interference mitigation techniques to enhance data extraction are
deemed beneficial to a successful UWB receiver design.
[0007] A UWB receiver design that utilizes transmitted reference
(TR) pulses is described in U.S. Patent No. 2001/0053175, entitled
"Ultra-Wideband Communications System," to Hoctor et al., patented
Dec. 20, 2001, including the following: "[t]he present invention
consists of the combination of two chief features and innovation
surrounding each of them. The first of these is known in the art as
transmitted-reference (TR). The TR technique is defined as the
transmission of two versions of a wideband carrier, one modulated
by data and the other unmodulated . . . The second feature of the
present invention is a type of multiple access scheme called `delay
hopping`. The term `delay hopping` refers to a multiple access
technique that is related to delay modulation in the way that
`frequency hopping` is related to frequency modulation . . . Our
new system has high immunity to narrow-band interference by virtue
of both the delay-hopping (DH) feature and the use of more than two
pulses in the TR transmission."
[0008] Conventional ultra wideband TR receivers perform poorly in
low-signal-to-noise-ratio environments due to an increase in the
received signal's noise-on-noise component introduced by various
types of channel noise, such as, AWGN or narrowband interference
(NBI) on the transmitted signal. Therefore, such receivers are
essentially ineffective for covert, low probability of intercept
and detection (LPI/D) communication systems where low power UWB
signals have to overcome the effect of intentional and
non-intentional jamming by narrowband signals. Moreover, since
conventional TR receivers use the same pulse shape for several
users in a multiple access channel, the correlation between similar
pulses increases the multiple access interference (MAI) as the
number of users increases.
[0009] Accordingly, a need exists for an ultra wideband
communication method and system that can suppress various types of
UWB and non-UWB interference prior to signal detection in a
multiple access communications format.
SUMMARY OF THE INVENTION
[0010] The present invention is directed to an ultra-wideband (UWB)
receiver, system, and method that utilizes a feedback loop
mechanism to enhance the signal-to-noise ratio of reference pulses
in a conventional TR receiver.
[0011] Another aspect of the present invention is directed to an
ultra-wideband (UWB) receiver that utilizes a feedback loop
mechanism to enhance the signal-to-noise ratio of reference pulses
in addition to correlating and multi-sampling received pulses over
values indicative of a second order statistical function from
received transmitted pulse pairs.
[0012] Another aspect of the present invention is directed to an
ultra-wideband (UWB) receiver that utilizes a feedback loop
mechanism to enhance the signal-to-noise ratio of reference pulses
in addition to correlating and multi-sampling received pulses over
values indicative of the autocorrelation function from received
transmitted pulse pairs.
[0013] Another aspect of the present invention is directed to an
ultra-wideband (UWB) system that utilizes a feedback loop mechanism
to enhance the signal-to-noise ratio of reference pulses in
addition to correlating and multi-sampling received pulses over
values indicative of the autocorrelation function to determine an
output signal that indicates encoded information.
[0014] Another aspect of the present invention is directed to an
ultra-wideband (UWB) method that includes: receiving one or more
UWB transmitted pulse pairs, each of the transmitted pulse pairs
separated by a predetermined lag interval D, wherein one or more
received pulse pairs have a predetermined encoded data and a
predetermined symbol repetition rate; feedback looping one or more
times, a predetermined portion of the one or more received UWB
pulse pairs, wherein the pulse pairs are shifted by the symbol
repetition period with each iteration through a feedback loop;
delaying a predetermined portion of the received one or more UWB
pulse pairs, by the lag interval D; multiplying and integrating one
or more reference signals produced by the looped pulse pairs with
one or more delayed versions of encoded data included in the
delayed one or more pulse pairs to generate a total energy that
decodes the one or more received pulse pairs; and outputting bit
information indicative of encoded information of the received pulse
pairs.
[0015] A final aspect of the present invention is directed to an
ultra-wideband (UWB) method of receiving one or more UWB
transmitted pulse pairs, each of the transmitted pulse pairs
separated by a predetermined primary lag interval D.sub.n, wherein
one or more received pulse pairs have a predetermined encoded data
and a predetermined symbol repetition rate; feedback looping one or
more times, a predetermined portion of the one or more received UWB
pulse pairs, wherein the received one or more pulse pairs are
shifted by the symbol repetition period with each iteration through
a feedback loop; delaying a predetermined portion of the received
one or more UWB pulse pairs, by the primary lag interval D.sub.n;
multiplying and integrating the primary and secondary delayed
pulses with a predetermined looped received reference pulse to
generate a plurality of values indicative of the autocorrelation
function of the received pulse pairs; matching a plurality of
stored sampled values indicative of the autocorrelation function of
the transmitted pulse pairs with the generated plurality of values
indicative of the autocorrelation function of the received pulse
pairs; and outputting bit information based on the matched sampled
values indicative of the autocorrelation function of the
transmitted pulse pairs that indicates encoded information of the
transmitted pulse pairs.
[0016] Accordingly, the present invention provides a desired UWB
system and method capable of delivering a large amount of data with
a low power spectral density by suppressing AWGN and NBI as well as
MAI. Such a system is more robust and has an increased signal to
noise coupled with added security for multiple users to communicate
in a covert communication system for military applications, such as
wireless sensor networks in a tactical battle field.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The accompanying drawings, which are incorporated into and
form a part of the disclosure, illustrate an embodiment of the
invention and, together with the description, serve to describe the
invention.
[0018] FIG. 1 is a general block diagram of a conventional UWB-TR
receiver in the presence of AWGN and multiple narrowband
interferers.
[0019] FIG. 2 shows the performance degradation of a conventional
TR receiver in an AWGN only channel compared against a transmission
channel that includes AWGN in addition to NBI.
[0020] FIG. 3 shows a modified TR receiver block diagram utilizing
a feed back loop for non-UWB interference mitigation.
[0021] FIG. 4 illustrates a reference pulse "Ref" enhancing example
after 3 iterations through an UWB receiver feedback loop.
[0022] FIG. 5(a) shows a TR doublet in a transmission channel that
includes AWGN and NBI.
[0023] FIG. 5(b) shows "Ref" pulse cleaning in a transmission
channel that includes AWGN and NBI.
[0024] FIG. 6 illustrates a comparison of Bit Error Rate (BER)
performance improvement for a TR receiver having a feedback loop
mechanism and a conventional TR receiver.
[0025] FIG. 7 shows a receiver embodiment utilizing a multi-pass
multiple delay arrangement (MPMD) coupled with a feedback loop.
DETAILED DESCRIPTION OF THE INVENTION
[0026] Referring now to the drawings, specific embodiments of the
invention are shown. The detailed description of the specific
embodiments, together with the general description of the
invention, serves to explain the principles of the invention.
[0027] Unless otherwise indicated, all numbers expressing
quantities of ingredients, constituents, reaction conditions and so
forth used in the specification and claims are to be understood as
being modified in all instances by the term "about". Accordingly,
unless indicated to the contrary, the numerical parameters set
forth in the specification and attached claims are approximations
that may vary depending upon the desired properties sought to be
obtained by the subject matter presented herein. At the very least,
and not as an attempt to limit the application of the doctrine of
equivalents to the scope of the claims, each numerical parameter
should at least be construed in light of the number of reported
significant digits and by applying ordinary rounding techniques.
Notwithstanding that the numerical ranges and parameters setting
forth the broad scope of the subject matter presented herein are
approximations, the numerical values set forth in the specific
examples are reported as precisely as possible. Any numerical
value, however, inherently contain certain errors necessarily
resulting from the standard deviation found in their respective
testing measurements.
General Description
[0028] Due to overlay of UWB and narrowband signals in a frequency
spectrum, the presence of narrowband interference in UWB
communication systems is often an unavoidable problem. Narrowband
interferers may have less energy than the UWB signal, but since
their energy is concentrated over a narrow bandwidth, they can mask
low power UWB signals. Although UWB communication systems have high
processing gain that can provide an inherent immunity to narrowband
interference, their performance can suffer considerably in the
presence of strong Narrow Band Interferers (NBIs). Therefore,
applying additional interference suppression techniques are
beneficial in a successful UWB receiver design.
[0029] The present invention thus provides a UWB communications TR
receiver that can operate in the presence of additive white
Gaussian noise (AWGN) and interference from coexisting narrowband
wireless systems by utilizing a feedback loop so as to suppress
such noise and improve the bit-error-rate performance of the
receiver and system disclosed herein.
[0030] Moreover, the present invention can be arranged to
additionally utilize a multiple pulse multiple delay (MPMD) method
and receiver design, which incorporates orthogonal TR modulated
chirp pulses for UWB channelization. Such a multiple pulse multiple
delay (MPMD) method uses time delays to separate the channels,
which can be as small as about a fraction of the pulse duration,
e.g., down to about a picosecond, for each user to produce the
improved BERs, i.e., the percentage of bits that have errors
relative to the total number of bits received in the transmission.
A similar MPMD structure and method thereof is disclosed in
incorporated by reference co-pending U.S. Application Ser. No.
60/462,507, titled "Multi-pulse Multi-delay (MPMD) Multiple Access
Modulation for UWB" by Dowla et al., the disclosure of which is
herein incorporated by reference in its entirety.
[0031] MPMD method uses orthogonal chirp pulses to modulate
transmitted data. The chirp pulses are typically generated by
conventional techniques, such as, voltage controlled oscillators
(VCO). The pulses are just a sequence of impulses; there is no
carrier. There is no specific frequency associated with this
technique; rather, its frequency spectrum is related by, for
example, the Fourier transform of the pulse. The free-space
radiated pulse may have a pulse-width of about 5 ns or less, often
about 1 ns or less, and more often about 200 picoseconds wide with
average power levels of up to about 5 mW. In addition, other pulse
shapes that provide orthogonality and decrease MAI and increase
channel capacity, such as, for example, a wavelet based waveform,
(i.e., a functional waveform that satisfies certain mathematical
requirements to represent other functions), or a Hermite based
waveform may also be employed.
[0032] A symbol in the present invention is represented by a pair
of pulses (unique to each user) called a reference ("Ref") pulse
and Data, i.e., "Tr" pulse separated by a unique delay for each
user. An exemplary UWB receiver of the present invention can
suppress interference for reference pulses in a TR receiver as
disclosed herein, by introducing a feedback loop with a delay equal
to a symbol repetition period (SRP) as applied to a conventional TR
receiver. In addition, such a feedback loop can be arranged in a
receiver design that can sample the receiver autocorrelation
function (ACF) at both zero and non-zero lags, thus also providing
a multi-sampling method that matches the shape of the received ACFs
rather than just the shape of the received pulses. Such a
multi-sampling, i.e., sampling zero and non-zero ACF lags, is
another beneficial approach when utilizing UWB pulses because of
channel distortions, such as, noise, fading and multipath effects,
(i.e., effects that result from the destructive interference caused
by the sum of several received paths that may be out of phase with
each other), which can make pulse detection difficult because the
pulse shape at the receiver is typically different compared to the
transmitted pulse shape. However, autocorrelation functions of
these pulses preserve their shape at the receiver despite the
deleterious effects of such channel distortions.
[0033] The present invention thus combines a feedback loop method
and system, orthogonal pulse shape coding, to conventional TR
receivers to suppress narrow band interferers (NBI) and additive
white Gaussian noise (AWGN), improve bit error rate (BER)
performance, reduce MAI, and increase channel capacity.
Specific Description
[0034] Turning now to the drawings, FIG. 1 shows a conventional
UWB-TR transceiver, generally designated as reference numeral 100,
which includes an UWB TR transmitter 12 configured to output UWB
signal(s), (s)t 14, in the presence of AWGN n(t) 16 and
interference from coexisting narrowband wireless systems, i.e., NBI
18. A conventional TR UWB receiver 22 (as shown by the dashed box)
as part of transceiver 100, is shown configured to receive signals
(shown as (r)t 20) from UWB TR transmitter 12 and such a receiver
22 includes a delay 24, denoted as D, split off a main line of the
received signal (r)t 20 that includes transmitted pulse pairs from
transmitter 12. Also include in conventional receiver 22 is a
multiplier 26 for multiplying the delayed and non-delayed versions
of transmitted pulse pairs encompassed within received signal (r)t
20, an integrator 28, and a hard decision block 30 that can output
bit information of the encoded data. Received signal (r)t 20 often
includes UWB signal (s)t 14, coupled with cumulative NBI 22 and
AWGN n(t) 16. The cumulative NBI 22 from K narrowband interferers,
having different carrier frequencies, k = 1 K .times. .times. m k
.function. ( t ) , ##EQU1## with AWGN n(t) 16 and UWB signal (s)t
14, as follows: r .function. ( t ) = s .function. ( t ) + k = 1 K
.times. .times. m k .function. ( t ) + n .function. ( t ) ( 1 )
.times. ( a ) ##EQU2##
[0035] The sinusoidal NBI 22 can be expressed in a simple form as:
m.sub.k(t)=M.sub.ksin(.omega..sub.kt) (k=1, . . . , K); (1)(b)
where .omega.=2.pi.f.sub.k.
[0036] The dominant component of interference caused by AWGN to TR
receivers in low SNR channels is identified as noise-on-noise
interference (I.sub.nn) defined by: I nn .function. ( AWGN ) =
.intg. T in .times. n .function. ( t ) n .function. ( t - D )
.times. .times. d t ##EQU3##
[0037] with its variance calculated as: E .times. { I nn 2
.function. ( AWGN ) } = E .times. { [ .intg. T in .times. n
.function. ( t ) n .function. ( t - D ) .times. .times. d t ] 2 } =
T in B N 0 2 2 . ( 3 ) ##EQU4##
[0038] Where T.sub.in represents the finite integration period, B
is the transmission bandwidth, and N.sub.0/2 corresponds to the
variance of AWGN. When NBI is added to a transmitted signal, the
value of I.sub.nn at the output of integrator in a TR receiver can
be calculated as shown by equations 4 and 5: I nn .function. ( NBI
) = .intg. T in .times. k = 1 K .times. .times. M k sin .function.
( .omega. k .times. t ) j = 1 K .times. .times. M j sin .function.
( .omega. j .function. ( t - D ) ) .times. .times. d t ( 4 ) I nn
.function. ( NBI ) = .intg. T in .times. k = 1 K .times. .times. j
= 1 K .times. .times. M k .times. M j 2 .times. cos .function. ( (
.omega. k + .omega. j ) .times. t - .omega. j .times. D ) cos
.function. ( ( .omega. k - .omega. j ) .times. t + .omega. j
.times. D ) .times. .times. d t ( 5 ) ##EQU5##
[0039] An integrator as shown by equation (5), represents a
low-pass filter, thus the low frequency sinusoids,
sin((.omega..sub.k-.omega..sub.j)t), and constant terms,
.omega..sub.jD, from the second sinusoidal term in (5) add up and
can cause significant distortion to a received signal. The variance
of such interference in a simple form of one interferer is given
by: E .times. { I nn 2 .function. ( NBI ) } = .times. E .times. { [
.intg. T in .times. M 2 sin .function. ( .omega. .times. .times. t
) sin .function. ( .omega. .times. .times. ( t - D ) ) .times.
.times. d t ] 2 } = .times. 1 4 .times. M 4 .times. T in 2 .times.
cos 2 .function. ( .omega. .times. .times. D ) ( 6 ) ##EQU6##
[0040] FIG. 2 shows the performance degradation of a conventional
TR receiver, as shown in FIG. 1, in an AWGN only channel 52 (shown
as plus symbols) compared against a transmission channel that
includes AWGN in addition to NBI 54 (shown as circles) from
multiple interferers. The x-axis shows the Signal to Noise Ratio
(SNR) in decibels (dB) that is an indication of the ratio of signal
power to noise power and the y-axis is the BER. As shown in FIG. 2,
the performance of a TR receiver in the presence of narrowband
interferers 54 is about two orders of magnitude worse than its
performance in an AWGN only channel 52.
[0041] FIG. 3 shows a receiver block diagram, generally designated
as reference numeral 300, of a novel receiver method of actively
suppressing interference in TR receivers while preserving a desired
UWB signal. In such a novel example embodiment, interference
suppression is achieved by introducing a feedback loop 312 to
enhance, received pulses (not shown) in TR doublets (i.e.,
reference "Ref" and data pulses "Tr" encompassed within received
signal r(t)), such as, but not limited to, Radio Frequency (RF)
pulse doublets. Feedback loop 312 is more often designed to have an
averaging delay, denoted by the letter T, equal to a symbol
repetition period (SRP), and a feedback loop 312 loss factor
.alpha. having a reduction factor of less than 1.
[0042] As illustrated in FIG. 3, an upper arm signal 314 is first
circulated in feedback loop 312 and back to an adder 313 instead of
feeding directly through to a multiplier 316 input, then through a
finite integrator 318 and finally to a decision block 320, wherein
T.sub.in, of finite integrator 318 denotes a predetermined finite
integration period of overlapped "Ref" and "Tr" pulses. A portion
of received signal (r)t is split off of a common input line 322 and
directed to a delay 324, denoted by D, which delays incoming pulse
pairs to enable the overlapping of "Ref", i.e., reference pulses,
captured in feedback loop 312 and "Tr" pulses encoded with data,
where their product as produced by multiplier 316 and integrated by
finite integrator 318, decodes the pulse pair by capturing the
total energy.
[0043] The overlapping of the "Ref" occurs when loop delay T is set
to the symbol repetition period (SRP). Since the polarity of "Ref"
pulses is always the same, the overlap in each loop iteration makes
such pulses gain in amplitude while data pulses "Tr" do not
experience the same resonance because of having opposite polarity
depending on the transmitted data.
[0044] Assuming that interference caused by AWGN is uncorrelated
with a desired signal, each circulation of input signal through
feedback loop 312 makes "Ref" pulses cleaner, i.e., a greater
signal to noise ratio, by rejecting the interference. However, in
order to have a successful narrowband interference rejection, the
feedback loop delay, T, should not be equal to integer multiplies
of interfering narrowband signal period to avoid resonating the
NBI.
[0045] FIG. 4, illustrates graphically, a reference pulse "Ref"
(shown as dashed pulses) increasing in signal after 3 iterations
through an UWB receiver feedback loop 412 of the present invention.
At stage A, an input signal containing data bits equal to {1,0,1}
is represented. At stage B, i.e., iterations 1-3, input pulses, are
shifted by a symbol repetition period, denoted as SRP, by feedback
loop 412 having an averaging delay T, resulting in a resonance in
overlapped "Ref" pulses prior to input to multiplier 416. At stage
C, an input symbol is delayed by D, through delay 424 which
represents the delay between the pulses in a UWB-TR doublet, and
overlaps a data pulse 434 with a strong clean reference pulse 438.
It is beneficial that feedback loop 412 amplifier implements a gain
of less than one to introduce a loss factor for processing gain of
the combined noise and signal associated with loop iterations. Such
adjustment of the gain results in noise signals circulating in the
loop along with the desired signals to incur an increment of loss.
This loss is often adjusted to the maximum value while assuring
loop stability. The lower the loop loss the longer would the loop
retain a pulse, hence the larger number of pulses that are
averaged.
[0046] FIG. 5(a) and FIG. 5(b) show "Ref" pulse cleaning in a
transmission channel that includes AWGN and NBI. Specifically FIG.
5(a) shows a plot of an UWB TR doublet, i.e., a reference "Ref"
pulse 502 and an encoded data "Tr" pulse 506, in the presence of a
strong interference 504. FIG. 5(b) shows a cleaned "Ref" pulse 508
after 100 loop iterations in the presence of reduced interference
510 having a loop gain of .alpha.=0.95. FIG. 5(b) also illustrates
"Tr" pulse 512 not exhibiting resonance after the same number of
loop iterations as "Ref" pulse 508.
[0047] Accordingly, as illustrated in FIG. 5(a) and FIG. 5(b), the
signal-to-interference ratio of "Ref" pulses increase significantly
as the number of loop iterations increases. Such a method is
beneficial for signals corrupted by AWGN channels, since different
samples of white noise are uncorrelated, although the feedback loop
filter often can introduce some correlation. Further, for a
successful NBI rejection, averaging delay T cannot be equal to
integer multiples of interfering narrowband signal period to avoid
resonating the NBI.
[0048] FIG. 6 illustrates a comparison of BER performance
improvement for a TR receiver of the present invention, as shown in
FIG. 3 having a feedback loop mechanism 652 (shown as circled
symbols) and a conventional TR receiver 654 (shown as plus symbols)
in a combined AWGN/NBI transmission channel.
[0049] FIG. 7 shows another example receiver embodiment of the
present invention, generally designated as reference numeral 700,
utilizing a multi-pass multiple delay arrangement MPMD coupled with
a feedback loop 702 having an averaging delay T and a gain .alpha.
of less than 1. As shown in FIG. 7, receiver 700 includes a user's
delay, i.e., a main delay (D.sub.n) 712, which represents the delay
between "Ref" and "Data" pulses in TR doublets for each user, and a
plurality of offset delays or lags (d.sub.nk), 716, 718, 720, of
the original received signal, denoted as r(t). An upper arm signal
704 is first circulated in feedback loop 702 instead of feeding
through directly to a multiplier 724 input. By utilizing multiple
delayed versions of received signal r(t) and multiplying and
integrating by a plurality of multiplier 724 and integrating
circuits 728 respectively with iterated "Ref" and "Data" pulses
through feedback loop 702, sampling of received signal's r(t)
second order statistical shape, such as, for example, over an
autocorrelation shape (and not on the signal shape), can be
performed at multiple times. Such a technique decreases multiple
access interference (MAI), increases channelization while
simultaneously suppressing undesired AWGN and NBI through the
feedback mechanism to further improve system BER performance.
[0050] As disclosed in incorporated by reference U.S. Application
Ser. No. 60/462,507, output signals from integrators 728 are
matched by a bank of matched filters 36 to sampled values of
autocorrelation functions (ACFs) of the original transmitted
pulses. The matched signals are then output, as denoted by the
letter Z, to a hard decision block 40 that can output a signal Y
indicative of the original encoded transmitted information.
[0051] Such an example embodiment, as shown in FIG. 7, more often
includes chirp pulses (i.e., a frequency modulated signal) with
different start and end frequencies with each user having its own
unique pulse shape. Chirp pulses that do not overlap in frequency
band and are theoretically uncorrelated with each other (i.e., are
orthogonal) can be separated using MPMD techniques. However, other
pulse formats, capable of providing uncorrelated pulse shapes,
known by those skilled in the art, which are suitable for UWB
communications can additionally be employed within the design
parameters of the present invention, such as, but not limited to,
Hermite function based orthogonal pulses and wavelet waveforms.
[0052] A typical UWB pulse can experience significant degradations
in dense multi-path channels but the shape of its second order
statistical function, e.g., the autocorrelation function (ACF), is
relatively preserved at the receiver. Therefore, multi-sampling the
shape of the second order statistical function by MPMD receiver 700
provides significant improvements to the detection process of
multiple pulse systems. As another arrangement, matched filters 736
can store cross-correlation functions of received signal r(t),
instead of stored ACFs of the original transmitted pulse(s), and
match these stored cross-correlation functions to cross-correlation
functions produced by similar circuitry of the present invention.
The matched signals are then output, as denoted by the letter Z, to
a hard decision block 40 that can output a signal Y indicative of
the original encoded transmitted information.
[0053] In a beneficial example method embodiment of the invention,
receiver 700 can be arranged to iterate through a feedback loop,
predetermined "Ref" and "Tr" pulses to suppress noise, such as, NBI
and AWGN, and in addition, multi-sample a second order statistical
function, such as, an ACF of each user's pulse at both the zero and
non-zero lags (i.e., ACF side lobes) and match them to
corresponding samples taken from known transmitted ACF pulse-shapes
rather than sampling and matching signal pulse shapes. Such an
example receiver 700 employs an autocorrelation technique by using
multiple delayed versions of the received signal. It is significant
to note that D.sub.n and d.sub.nk are unique for each receiver
channel, i.e., each user. Then, matched filters 736 matched to ACF
samples, denoted as .lamda..sub.k.sup.(n) in FIG. 7, of transmitted
pulses for each user followed by hard decision block 740 separates
the channels. The output of receiver 700, as denoted by the letter
Y in FIG. 7, is given by: Y=(sgn(.lamda..sub.k.sup.(n){circumflex
over (r)}.sub.k,m.sup.(n)) Here, the dot product represents matched
filtering, n=1, 2, N (# of users), k=1, 2 , .times. , K ( .English
Pound. .times. .times. of .times. .times. lags ) , m = 1 , 2 ,
.times. , M ( .English Pound. .times. .times. of .times. .times.
bits ) , and .times. .times. where , .times. .lamda. k ( n ) = [ R
P n .times. P n ( n ) .function. ( 1 ) .times. .times. .times.
.times. R P n .times. P n ( n ) .function. ( K ) ] ; .times. and
##EQU7## r ^ k , m ( n ) = [ r ^ 1 , 1 ( n ) r ^ 1 , m ( n ) r ^ k
, 1 ( n ) r ^ k , m ( n ) ] ; ##EQU7.2## wherein .lamda..sub.k(n)
denotes a vector of K sampled values of the n.sup.th user's
transmitted pulses' normalized autocorrelation function. For
example, R.sub.P.sub.n.sub.P.sub.n.sup.(n)(1) represents the first
sampling point (lag=0) of the normalized ACF for the n.sup.th
user's transmitted pulse. Similarly, {circumflex over
(r)}.sub.k,m.sup.(n), shown as the output of integrators 728 in
FIG. 7, provides the sampled ACF matrix of the n.sup.th user's
m.sup.th bit for k lags at the receiver. Each row of the
{circumflex over (r)}.sub.k,m.sup.(n) matrix represents the
correlation between the received signal and its delayed version
based on various delays as shown by the following equation: r ^ k ,
m ( n ) = .intg. ( m - 1 ) .times. T ( m - 1 ) .times. T + T in
.times. r .function. ( t ) . r .function. ( t - .DELTA. n ) .times.
.times. d t ; ##EQU8## where T.sub.in is the integration time that
is most effective at a pulse width and .DELTA..sub.n is the total
delay given by: .DELTA..sub.n=D.sub.n+d.sub.nk.
[0054] As shown by the equation for .DELTA..sub.n, D.sub.n
represents the n.sup.th user's main delay representing lag zero in
the ACF and d.sub.nk denotes the offset from its main delay or lag
k in the ACF and its value is zero for k=1 (d.sub.n1=0).
[0055] Therefore, delaying the received signal by D.sub.n causes a
"Ref" pulse to align with a "Tr" pulse in each pulse pair where
their product decodes the pulse pair by capturing the energy in lag
zero of the generated ACF. Integrating this product over a finite
time samples the ACF at lag zero. Further delaying the received
signal by multiple offsets d.sub.nk, which is added to the main
delay D.sub.n and multiplying with its undelayed version, samples
the ACF in non-zero lags after integration. The sampled points are
estimates of the received signal's ACF. These values ({circumflex
over (r)}.sub.k,m.sup.(n)) are then matched to the original pulses'
sampled ACF values (.lamda..sub.k.sup.(n)) and provide a more
accurate decoding of the received symbols.
[0056] It should be understood that the invention is not intended
to be limited to the particular forms disclosed. Rather, the
invention is to cover all modifications, equivalents, and
alternatives falling within the spirit and scope of the invention
as defined by the following appended claims.
* * * * *