U.S. patent application number 14/016273 was filed with the patent office on 2015-03-05 for method for symbol sampling in a high time delay spread interference environment.
This patent application is currently assigned to HARRIS CORPORATION. The applicant listed for this patent is HARRIS CORPORATION. Invention is credited to MAC L. HARTLESS, Richard D. Taylor, Steve R. Wynn.
Application Number | 20150063512 14/016273 |
Document ID | / |
Family ID | 52443670 |
Filed Date | 2015-03-05 |
United States Patent
Application |
20150063512 |
Kind Code |
A1 |
HARTLESS; MAC L. ; et
al. |
March 5, 2015 |
METHOD FOR SYMBOL SAMPLING IN A HIGH TIME DELAY SPREAD INTERFERENCE
ENVIRONMENT
Abstract
Symbol sampling in a high time delay spread interference
environment includes acquiring (602) a time varying baseband
waveform. The waveform has a signal amplitude that varies between
one of a plurality of symbol states. The waveform is sampled (603)
at a rate of m times the symbol rate. During an evaluation time, an
error value is calculated (604, 606) for each of m data sample
positions. Each of the error values comprises an average distance
between the measured value of the waveform as indicated by the data
sample and a closest known symbol value. The error values are used
to create an error surface. Thereafter, the error surface is
modeled as a quadratic and an optimal sample time is determined
(608, 610, 612) based on finding the time location where the
quadratic surface is minimum. A sinc interpolator is then used to
resample the data.
Inventors: |
HARTLESS; MAC L.; (Forest,
VA) ; Taylor; Richard D.; (Moneta, VA) ; Wynn;
Steve R.; (Lynchburg, VA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
HARRIS CORPORATION |
Melbourne |
FL |
US |
|
|
Assignee: |
HARRIS CORPORATION
Melbourne
FL
|
Family ID: |
52443670 |
Appl. No.: |
14/016273 |
Filed: |
September 3, 2013 |
Current U.S.
Class: |
375/355 |
Current CPC
Class: |
H04L 7/0062 20130101;
H04L 7/0029 20130101 |
Class at
Publication: |
375/355 |
International
Class: |
H04L 7/00 20060101
H04L007/00 |
Claims
1. A method for symbol sampling in a digital communication system,
comprising: acquiring a time varying baseband waveform in which a
signal amplitude varies between one of a plurality of symbols;
sampling the waveform at a rate of m times the symbol rate; during
an evaluation time, calculating an error value for each of m data
sample positions, each said error value comprising an average
distance between the measured value of said waveform as indicated
by said data sample and a closest known symbol value; and
determining an optimal sample time based on said plurality of error
values.
2. The method according to claim 1, wherein said determining of
said optimal sample time further comprises identifying an
approximate time relative to said data sample positions when said
error value is minimized.
3. The method according to claim 1, wherein said determining of
said optimal sample time further comprises selecting at least one
of said m data sample positions where said error value is
minimized.
4. The method according to claim 3, wherein a plurality of said m
data sample positions are selected and said determining of said
optimal sample time further comprises calculating a curve which is
fitted to the filtered error values corresponding to the selected
ones surrounding the data sample position where said error value is
minimized.
5. The method according to claim 4, wherein an inflection point of
the curve defines an optimal sample time when said error value is
minimized.
6. The method according to claim 4, wherein said curve is defined
by a quadratic equation.
7. The method according to claim 1, further comprising
interpolating to determine an estimated sample value of said
waveform corresponding to a sample time which is intermediate of
said m data sample positions.
8. The method according to claim 1, further comprising generating a
resample filter based on said optimal sample time.
9. The method according to claim 8, further comprising re-sampling
said time varying baseband waveform using said resample filter to
estimate said signal amplitude at said optimal sample time.
10. The method according to claim 1, further comprising selecting
said evaluation time to be of a minimum duration sufficient to
allow convergence to an optimal sample time.
11. A radio frequency (RF) communication device, comprising an RF
receiver configured to acquire a time varying baseband waveform in
which a signal amplitude varies between one of a plurality of
symbol states; and at least one electronic circuit which is
configured to sample the waveform at a rate of m times the symbol
rate; during an evaluation time, calculate an error value for each
of m data sample positions, each said error value comprising an
average distance between the measured value of said waveform as
indicated by said data sample and a closest known symbol value; and
determine an optimal sample time based on said plurality of error
values.
12. The RF communication device according to claim 11, wherein said
at least one electronic circuit is further configured to determine
said optimal sample time by identifying an approximate time
relative to said data sample positions when said error value is
minimized.
13. The RF communication device according to claim 11, wherein said
at least one electronic circuit is further configured to determine
said optimal sample time by selecting at least one of said m data
sample positions where said error value is minimized.
14. The RF communication device according to claim 13, wherein said
at least one electronic circuit is further configured to select a
plurality of said m data sample positions, and determines said
optimal sample time by calculating a curve which is fitted to the
error values corresponding to the surrounding ones of the data
sample position where said error value is minimized.
15. The RF communication device according to claim 14, wherein an
inflection point of the curve defines an optimal sample time when
said error value is minimized.
16. The RF communication device according to claim 14, wherein said
curve is defined by a quadratic equation.
17. The RF communication device according to claim 11, wherein said
at least one electronic circuit is further configured to
interpolate an estimated sample value of said waveform
corresponding to a sample time which is intermediate of said m data
sample positions.
18. The RF communication device according to claim 11, wherein said
at least one electronic circuit is further configured to generate a
resample filter based on said optimal sample time.
19. The RF communication device according to claim 18, wherein said
at least one electronic circuit is further configured to re-sample
said time varying baseband waveform using said resample filter to
estimate said signal amplitude at said optimal sample time.
20. The RF communication device according to claim 11, wherein said
evaluation time is of a minimum duration sufficient to allow
convergence to an optimal sample time.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Statement of the Technical Field
[0002] The inventive arrangements relate to radio communications
and more particularly to improvements in symbol sampling in
communications environments where a significant amount of delay
spread exists.
[0003] 2. Description of the Related Art
[0004] Simulcast systems create delay spread at receive sites that
is proportional to site spacing. The delay spread is known to
degrade the bit-error-rate (BER) of receive radios in site overlap
areas. Coverage area of a simulcast system can be improved by
reducing the adverse effects that delay spread has upon BER. When
the coverage area is improved in this way, the number of simulcast
transmitter sites can be reduced, thereby lowering overall
infrastructure costs.
[0005] Symbol synchronization is used to estimate the optimal time
for sampling each symbol in a received signal. Improved symbol
synchronization is a key factor for purposes of reducing BER.
Conventional synchronization methods involve correlating a burst of
a limited number of received symbols contained in a
"synchronization field" with a known set of reference symbols. The
symbols contained in the synchronization field do not represent
user data and therefore comprise overhead in the communication
system. Accordingly, the number of symbols contained in the
synchronization field and the frequency of such fields is
intentionally limited.
[0006] The foregoing synchronization approach maximizes the
available bandwidth for throughput of user data. Still, due to the
limited number of synchronization symbols, conventional
synchronization methods take a relatively long time to converge to
the optimal synchronization time when operating in a high delay
spread environment. This slow synchronization process can severely
degrade the time required to establish initial synchronization and
call start.
SUMMARY OF THE INVENTION
[0007] Embodiments of the invention concern a method for symbol
sampling in a digital communication system. The method involves
acquiring a time varying baseband waveform in which a signal
amplitude varies between one of a plurality of symbol states. The
waveform is sampled at m times the symbol rate, where m will be
referred to as the oversample rate i.e. there are m samples
covering a symbol period. During an evaluation time, an error value
is calculated for each of m data sample positions. Each of the
error values comprises an average distance between the measured
value of the waveform as indicated by the data sample and the
closest known symbol value. Thereafter, an optimal sample time is
determined based on the calculated error values.
[0008] According to another aspect, the invention concerns a radio
frequency (RF) communication device. The radio frequency
communication device includes an RF receiver. The RF receiver is
configured to acquire a time varying baseband waveform as described
above. The RF communication device also includes at least one
electronic circuit. The at least one electronic circuit is
configured to oversample the waveform at an oversample rate of m.
During an evaluation time, the at least one electronic circuit
calculates an error value for each of m data sample positions. Each
error value comprises an average distance between the measured
value of the waveform as indicated by the data sample and a closest
known symbol value. Subsequently, the at least one electronic
circuit determines an optimal sample time based on the calculated
error values.
BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Embodiments will be described with reference to the
following drawing figures, in which like numerals represent like
items throughout the figures, and in which:
[0010] FIG. 1 is a diagram that is useful for understanding the
concept of delay spread.
[0011] FIG. 2 is a simplified model that is useful for
understanding the effects of delay spread on an eye-pattern of
received signals.
[0012] FIG. 3 is a data structure diagram that illustrates a
traffic channel structure in a Project 25, Phase 2 compliant
system.
[0013] FIGS. 4A-4C are useful for understanding how an error
distance is determined using measured soft data and the nearest
known symbol values.
[0014] FIG. 5 is a schematic illustration of an exemplary
communication system that is useful for understanding the present
invention
[0015] FIG. 6 is a flow chart that is useful for understanding a
method for improving symbol sampling.
[0016] FIG. 7A is an exemplary eye pattern associated with a
received signal in a low delay spread environment.
[0017] FIG. 7B is a plot of the mean of the absolute value of the
error between the measured and the known symbol values for the
exemplary eye pattern in FIG. 4A.
[0018] FIG. 8A is an exemplary eye pattern associated with a
received signal in a high delay spread environment.
[0019] FIG. 8B is a plot of the mean of the absolute value of the
error between the measured and the known symbol values for the
exemplary eye pattern in FIG. 5A.
[0020] FIG. 9A shows an exemplary eye pattern plot.
[0021] FIG. 9B shows an exemplary plot of the mean of the absolute
value of the error between the measured and the known symbol values
for the exemplary eye pattern in FIG. 8A, where the optimal sample
position is close to one of the sampled positions.
[0022] FIG. 10A shows an exemplary eye pattern plot.
[0023] FIG. 10B shows an exemplary plot of the mean of the absolute
value of the error between the measured and the known symbol values
for the exemplary eye pattern in FIG. 9A, where the optimal sample
position is not centered on one of the sampled positions.
[0024] FIG. 11 is a drawing that is useful for understanding how an
error surface which is modeled as a quadratic equation refines the
sample position to a higher accuracy.
[0025] FIG. 12 is a Hann windowed Sinc function with optimal time
offset, which is used as a resample filter to interpolate a best
data value.
[0026] FIG. 13A is an eye pattern plot which shows data sampled
before interpolation.
[0027] FIG. 13B is an eye pattern plot which shows the data sampled
after interpolation.
[0028] FIG. 14 is an architectural diagram of a radio communication
system that is useful for understanding the invention.
DETAILED DESCRIPTION
[0029] The invention is described with reference to the attached
figures. The figures are not drawn to scale and they are provided
merely to illustrate the invention. Several aspects of the
invention are described below with reference to example
applications for illustration. It should be understood that
numerous specific details, relationships, and methods are set forth
to provide a full understanding of the invention. One having
ordinary skill in the relevant art, however, will readily recognize
that the invention can be practiced without one or more of the
specific details or with other methods. In other instances,
well-known structures or operation are not shown in detail to avoid
obscuring the invention. The invention is not limited by the
illustrated ordering of acts or events, as some acts may occur in
different orders and/or concurrently with other acts or events.
Furthermore, not all illustrated acts or events are required to
implement a methodology in accordance with the invention.
[0030] Referring now to FIG. 1 there is shown an exemplary
simulcast communication system 100. In system 100 the same data is
synchronously transmitted from two or more transmitter sites 101,
102 on the same frequency. The transmitter sites are situated in
different locations so that a larger coverage area is obtained as
compared to that which would be possible with only a single
transmitter. When a receiver 103 is located in an area that is
intermediate of the two or more transmitter sites, the received
signal will be a composite of all transmitters within range of the
receiver. Within this overlap region delay spread is experienced by
the receiver due to the different amounts of time required for
transmission of signals from each transmitter. For example, in the
example, shown the delay spread will be t.sub.a-t.sub.b, where
t.sub.a is the propagation delay from transmitter site 101 to
receiver 103, and t.sub.b is the propagation delay from transmitter
102 to receiver 103.
[0031] In a telecommunication system, a received digital data
signal can be repetitively sampled and displayed on an
oscilloscope. For certain types of data modulation schemes, the
resulting pattern that is displayed will look like a series of
eyes. For example, an exemplary eye pattern including eye 702 is
shown and discussed in relation to FIG. 7A. The eye pattern
represents the synchronized superposition of all realizations of
the signal of interest as viewed within a particular time period.
As is well known in the art, the eye pattern is a useful tool for
analyzing the effects of inter-symbol interference such as may be
caused by time delay spread.
[0032] Referring now to FIG. 2 there is shown a simplified model of
an eye pattern that is useful for understanding the effects of the
delay spread in FIG. 1. The model shows a 1st path eye 201 which
represents a symbol received at a first time t1. In this model,
delay spread is present and as a result, the same symbol is
received from a different transmitter at a time t2. This signal is
represented in the model as a 2nd path eye 202. The delay spread in
this example is t1-t2. In a scenario such as the one shown in FIG.
2, the usable time period for sampling of the received data occurs
during the region represented by the cross-hatching. This usable
time period is referred to herein as the usable eye 203. The
optimum time for sampling the usable eye in such a scenario is at
the center of the usable eye as shown. As delay spread increases,
the usable eye will decrease in duration and therefore the
importance of accurate symbol synchronization is increased.
[0033] In a digital radio communication system a receiver is time
synchronized with a transmitter to facilitate decoding of
transmitted symbols. Current terminal design methodology typically
involves correlating a known reference symbol pattern against a
burst of synchronization symbols contained in a synchronization
field of a received data transmission. However, the number of
synchronization symbols can be so limited within the data stream
that it takes a long time for the receiver to converge to optimal
time synchronization. In order to more fully understand this
problem, it is useful to consider the P25 standard for radio
communication.
[0034] The P25 Phase 2 Media Access Control (MAC) standard
describes a traffic channel which can be used for transmitting
voice data between two or more radio transceivers. A detailed
discussion of the P25 MAC standard is beyond the scope of this
invention. Briefly however, the standard describes a traffic
channel comprised of a plurality of ultraframes where each
ultraframe contains four superframes. The traffic channel is
logically subdivided into two voice channels which are designated
VCH0 and VCH1. The time slot structure of each superframe is
dependent upon whether the particular superframe is being used for
the inbound interface or the outbound interface.
[0035] Referring now to FIG. 3 there is shown a plurality of time
slots 305, 308 associated respectively with an outbound VCH0 voice
channel 302 and an inbound VCH0 voice channel 304 in accordance
with a P25 MAC standard. The outbound interface includes time slots
305, 308 contain voice information or other types of user data. A
signaling field is provided between time slots. The signaling field
is referred to as the Interslot Signaling Channel or ISCH. The ISCH
alternately includes two 40 bit segments which comprise either
information (I-ISCH) or a synchronization pattern (S-ISCH). The
S-ISCH signaling fields are identified as 306 in FIG. 3. The S-ISCH
fields are 20 symbol patterns (3.3 milliseconds in duration)
occurring on average, every 60 milliseconds. This synchronization
pattern or S-ISCH is used by a receiver to synchronize the decoding
of received symbols. The important thing to appreciate about the
S-ISCH for purposes of the present invention is that they comprise
a relatively small number of synchronization symbols as compared to
the transmitted data stream. Since the synchronization symbols are
so limited in quantity, the current method for determining time
synchronization takes a long time to converge to the center of the
usable eye when operating in a high delay spread environment. This
time delay substantially increases the time to establish initial
synchronization and begin a call.
[0036] As noted above, synchronization involves finding the optimal
sample position at the center of the usable eye. One aspect of the
present invention concerns methods and systems for accelerating the
rate at which a terminal can establish synchronization based on a
received digital data signal. Unlike systems that rely upon a
limited number of symbols contained in synchronization fields, the
arrangements described herein will use all received symbol data.
The method described herein will also compare measured symbol
values to known or expected symbol values. The result is a much
more rapid process for determining the optimal sample time.
[0037] In the process described herein, the received demodulated
soft data is first compared at each sample position to the k known
hard symbol values. The goal of this step is to find the distance
between the soft value and the nearest hard symbol neighbor. As
used herein, the phrase "soft data" is intended to refer to the
actual sampled or measured output produced by demodulating and
filtering a phase modulated communication signal such as a PSK,
C4FM or CQPSK type signal. When demodulated and filtered, such a
signal will typically have an amplitude that varies over time in
accordance with a plurality of symbol states. Each symbol state
will correspond to a have known or expected amplitude, and each
amplitude value has a predefined meaning associated with a specific
bit pattern. The known or expected amplitude for a particular
symbol state is referred to herein as the "hard symbol value".
[0038] FIGS. 4A and 4B show two examples where demodulated soft
data x.sub.1, x.sub.2 are evaluated to determine the distance
between measured soft data and the closest known hard symbol
values. In each case, the error is the distance between a measured
soft data value x.sub.i and the known hard symbol value of the
closest symbol Sk. In the examples shown k=0, 1, 2, 3 so there are
four possible known hard symbol values Sk which are represented in
the figures by values S0, S1, S2, and S3. In FIG. 4A the actual
measured soft data value for this particular sample is x.sub.1. The
closest hard symbol value neighbor in this example is S3. However,
the value of x.sub.1 is slightly less than the known value of S3 by
a distance or amount equal to d1. The distance d1 is the error in
this example. In FIG. 4B, the measured soft data value is x.sub.2.
The closest hard symbol neighbor in this example is S1. However,
the value of x.sub.2 exceeds the known hard symbol value of S1 by a
distance d2. The distance d2 is the error in FIG. 4B.
[0039] FIG. 4C is similar to FIGS. 4A and 4B but shows soft value
data is acquired (oversampled) at several sample times over the
duration of one symbol. Also, a usable eye 403 is superimposed over
the diagram to illustrate a time when each soft value data sample
x.sub.3, x.sub.4 and x.sub.5 is acquired relative to the timing of
the usable eye. It can be observed that soft value x.sub.4 is
acquired at approximately the center of the usable eye, whereas
soft values x.sub.3 and x.sub.4 are acquired nearer the extremities
of the usable eye. In general, error distance d will be minimized
when the sample time or position from which the soft data is
obtained is close to the center of the usable eye 403. It can be
observed in FIG. 4C that the error distance d4 is less than the
error distance d3 and d5. Accordingly, the sample position that
results in the minimum error distance, averaged over a number of
symbols will be close to the center of the usable eye. By
evaluating error distance in this way, an approximate center of the
usable eye 404 can be determined.
[0040] Once the approximate center of the usable eye is determined,
the process proceeds to a second step. In the second step, further
refinement is obtained regarding the optimal sample time associated
with minimum error. In this second step, the optimal sample time or
position is obtained via a curve fitting process that uses the
minimum average distance sample (e.g. soft data sample x.sub.4),
and the average distance computed for samples on each side of the
minimum distance sample (e.g. soft value samples x.sub.3 and
x.sub.5). A quadratic equation fitting process has been found to
work well for this step, but other curve fitting processes can also
be used. This fitting process facilitates a better estimation of a
minimum point in the error curve representing the optimal sample
time. The optimal sample time as determined by this second step is
then used to design a sinc interpolator to resample the original
data to the optimal location, effectively providing a higher symbol
oversample rate. The various aspects of the inventive arrangements
will now be described in further detail.
[0041] Referring now to FIG. 5 there is shown an exemplary
architecture for a digital data receiver including RF amplifier
502, mixer 504, demodulator 506, filter/equalizer 508, symbol
decoder 510, speech decoder 512, and D/A converter 514. The digital
data receiver can be configured to receive signals transmitted
according to a phase shift type of digital data modulation scheme
such as PSK, C4FM or CQPSK. Most of the components of the digital
data receiver are well known in the art and therefore will not be
described in detail. However, it will be appreciated that signals
received at antenna 501 are amplified in low noise RF amplifier 502
before being passed to mixer 502. The mixer uses a local oscillator
signal received from a local oscillator 514 to down-convert the
received RF signal to an intermediate frequency (IF). After
down-conversion, the IF signal is communicated to a demodulator 506
where the IF signal is demodulated to obtain a time varying
baseband signal in which the amplitude of the signal varies in
accordance with each symbol that has been received. After the data
has been demodulated at 506 additional filtering can be applied at
filter/equalizer 508. The resulting demodulated output, which has
an amplitude that varies over time in accordance with a received
symbol, is then communicated to a symbol decoder 510 to extract the
digital data bits represented by each symbol. The extracted data in
this example is speech information and so the data is communicated
to a speech decoder 512. The speech decoder converts the digital
data into digital audio data and passes this information on to the
D/A converter 514. The D/A converter will convert the digital audio
data into analog audio which can be reproduced at a loudspeaker
(not shown).
[0042] In the receiver system shown in FIG. 5, the symbol decoder
510 will periodically sample the amplitude or value of the
demodulated signal to determine what symbol has been sent. To this
end, it is advantageous for the symbol decoder 510 to have timing
information so as to determine the optimal moment when each symbol
should be sampled. The inventive arrangements facilitate
improvements in the operation of the symbol decoder by accelerating
the rate at which the symbol decoder can determine the optimal time
for sampling the symbol data. This process will now be described in
further detail in relation to FIG. 6.
[0043] A method for improving symbol sampling in a high time delay
spread interference environment begins at 601 and continues to 602
where a time varying baseband waveform is acquired. The baseband
waveform can be acquired using suitable receiver processing
circuitry similar to that described herein with respect to FIG. 5.
At 603, the baseband waveform is sampled at a rate m times the
symbol rate. For example the sample rate can exceed the symbol rate
by a factor of 3.times. or more. Note that a minimum value of m=3
is necessary to perform the quadratic fit operation. According to
one aspect of the invention, the sample rate can be 4.times.
greater than the symbol rate. At 604 an error metric err(t, m) is
calculated at the oversample rate, where t is a unit of time and m
is the oversample rate. The error determination technique was
generally described above in relation to FIGS. 4A-4C and can be
expressed mathematically as:
err(t,m)=Min over all k{abs(x(t,m)-sym(k))} (1)
where: x is the value of the received demodulated soft data t is a
sample time; m is the symbol oversample rate; k is an integer that
is less than or equal to the number of symbol state values; and
sym(k) is the known symbol value.
[0044] After the error err(t, m) has been calculated at 604, the
process continues on to 606 where the calculated value of err(t, m)
for each sample location m is filtered or averaged to obtain
Ferr(t, m). The values are averaged over some predetermined period
of time so as to obtain for each value of m a mean or average value
of err(t, m). At 607, the average error value at each sample
position m is used to plot an error surface.
[0045] In order to understand steps 602-607 there is shown in FIGS.
7A and 7B an exemplary eye pattern corresponding to the output of
filter/equalizer 508. The particular eye pattern in FIG. 7A is
associated with a low delay spread environment. Due to the fact
that the plot was obtained in a low delay spread environment, the
eye pattern is well defined and shows only minimal variation in the
pattern over time. In contrast, FIG. 8A shows a similar type
pattern acquired in a high delay spread environment (delay
spread=50 microseconds). The eye pattern in FIG. 8A is poorly
defined due to the delay spread of the received signals.
[0046] Referring now to FIG. 7B, there is provided an error curve
plotted from filtered or averaged error data captured from signals
associated with the eye pattern of FIG. 7A. The average values of
Ferr(t, m) at each sample time m are denoted in FIG. 7B by the
circular marks denoted on the plot. As such, the error curve shows
for each sample time m the mean of the absolute value of the error
as between the measured and known symbol values. It can be observed
in FIG. 7B that the error curve reaches a minimum when the time is
approximately 165 microseconds. It can also be observed that this
time corresponds closely to the center of the eye pattern in FIG.
7A. Similarly, FIG. 8B shows an error curve calculated for the data
captured in the high delay spread environment of FIG. 8A. Here, the
error metric reaches a minimum when the time is approximately 180
microseconds. In each example the error curve gives an approximate
indication of the center of the usable eye.
[0047] After the error curve has been generated, the process
continues on to step 608 where a sample time t on the error curve
is selected where the curve indicates minimum error. This step can
be expressed as: Loc(t)=min{Ferr(t,m) over all m}, where Loc(t) is
the time when the curve reaches a minimum value. For example, in
FIG. 7B this minimum error time would occur where t=165
microseconds. In FIG. 8B, the minimum error point is at
approximately t=185 microseconds.
[0048] In some instances, the minimum average error position
expressed as Loc(t) will be close to one of the positions where the
data has actually been sampled. This concept is illustrated in
FIGS. 9A and 9B wherein there is shown another example of an eye
pattern and a resulting error curve generated as described herein.
For convenience with regard to this explanation the time axis in
FIG. 9B is indexed in accordance with sample position values rather
than actual time values. It can be observed in FIG. 9B that the
error curve has a minimum mean value that appears to correspond
almost exactly with sample position 5. In contrast, with respect to
FIGS. 10A and 10B it can be observed that the amplitude of the
averaged absolute error is approximately the same magnitude at
sample positions 4 and 5. This suggests that the actual minimum
error is likely somewhere between these two sample points.
Accordingly, it is somewhat unclear in FIG. 10B exactly where the
optimum sample point should be. In order to address this problem
the process in FIG. 6 advantageously continues on to step 610
wherein the error surface (e.g. the error surface shown in FIG.
10B) is modeled to find a curve of best fit. In other words,
selected data points adjacent or close to the minima are used to
construct a smoothed error function that best explains the observed
data points. A quadratic fit of error surface has been found to
work well for this purpose. The fitted curve can be constructed for
Ferrr(t, m) over the domain m=Loc(t-1): Loc(t+1), where t=0 is the
sample time having the lowest mean error. Stated differently, it
can be said that we evaluate Ferr(t, Loc(t-1):Loc(t+1)) around the
lowest mean error value to construct a curve that is fitted to the
error data.
[0049] This foregoing concept is illustrated in FIG. 11 which shows
sample data y(-1), y(0) and y(+1) as measured at sample positions
-1, 0, and +1 (i.e., sample positions corresponding to Loc(t-1),
Loc(t) and Loc(t+1). Note that the error curve 1102 has a limited
number of data sample points and is therefore comprised of a
plurality of linear segments. The data points y(-1) and y(0) are at
approximately the same value, suggesting that the actual minimum
value may occur somewhere between sample time t=-1 and t=0. In
order to estimate the time t(opt) corresponding to the actual error
minimum y(min), a quadratic curve fitting process is used. The
standard form of a quadratic equation is:
y(t)=at.sup.2+bt+c (2)
By using the measured data values y(-1), y(0) and y(+1), the
constants a, b and c from equation (2) can be calculated as
follows:
[ 1 - 1 1 0 0 1 1 1 1 ] [ a b c ] = [ y ( - 1 ) y ( 0 ) y ( 1 ) ]
##EQU00001##
The inflection point of a best fit quadratic equation will provide
a useful estimate of the true minimum point of the error curve in
FIG. 11. A quadratic fit 1104 of the error surface 1102 is shown in
FIG. 11. The inflection point for a curve defined by a quadratic
equation (2) can be stated as:
t(opt)=-b/2a, (3)
where t(opt) represents an improved estimate of the optimum sample
time where the error curve will be at its minimum point y(min) as
suggested by the measured data and the quadratic curve fitting in
step 610. It can be observed in FIG. 11 that by using the foregoing
equation (3) the optimal sample time is determined to occur at a
normalized offset of approximately -0.45 relative to sample time
zero.
[0050] The foregoing steps 604-610 are useful for identifying the
optimum time when the received symbol data should have been
measured in order to minimize error. In most instances however, the
actual sample times will not coincide with the optimum sample time.
Accordingly, an interpolation step 612 is provided to estimate a
symbol value at t(opt). The symbol estimate will be an interpolated
symbol value based on the actual sample values obtained at sample
times before and after the optimum sample time. The goal here is to
compute a symbol value at an arbitrary continuous time t(opt) using
the available set of discrete-time samples. Any suitable
interpolation method can be used for this purpose. However, it has
been determined that a Hann windowed sinc function with a time
offset t(opt) can provide a resample filter which will effectively
interpolate a higher fidelity data value. Sinc interpolators are
well known in the art and therefore will not be described here in
detail. It should be noted that a windowed sinc interpolator is
preferred for its ability to maintain the fidelity of the resampled
signal. The length of the sinc filter which is applied can be
designed to maintain any desired level of interpolation loss.
[0051] As shown in FIG. 12, once the time offset value, t(opt), has
been found via finding the time location of the minimum of the
quadratic surface, a sinc filter can be designed to effectively
provide a sample of the input data at the time shifted location. In
FIG. 12 the sinc filter representing a time shift of -0.45 samples
corresponding to t(opt)=-0.45 is shown. The interpolation described
herein effectively provides a higher sample rate of the data, but
advantageously only needs to be performed once per symbol.
[0052] Referring now to FIG. 13A an eye pattern is shown wherein
data is sampled prior to interpolation. More particularly, it can
be observed that sample times 1302 do not align with the center of
each eye. Conversely, FIG. 13B shows that by using the methods
described herein the effective sample time 1304 occurs in alignment
with the center of the eye. Note here that we refer to an
"effective" sample time 1304 because the actual sample time (i.e.,
the time when the actual sampling is performed by the decoder
hardware) will not generally correspond to the effective sample
time after interpolation. In this regard, the interpolation
effectively shifts the sample time to the optimal sample time, and
provides the benefit of a higher sample rate. This higher effective
sample rate is accomplished while avoiding additional processing
and computational complexity normally required of such higher
sample rate. At 613 a determination can be made as to whether
additional data is to be acquired, and if so (613: Yes), then the
process returns to 602 and continues. Otherwise the process
terminates at 614.
[0053] The method described herein with respect to FIG. 6 provides
a significant advantage insofar as it converges to best sample
position much more rapidly as compared to conventional methods. For
example, in the context of a P25 waveform where delay spread is 60
microseconds, the algorithm described herein will converge to the
optimal sample time in less than 100 milliseconds, using all
received data. In contrast, conventional algorithms relying on the
limited number of synchronizing symbols for determining the optimal
sample position would require up to 1.8 seconds to arrive at the
optimal sample time. For purposes of the present disclosure, the
process will be deemed to have sufficiently converged to an optimal
sample time when further evaluation of the data will provide only
minimal improvement in the bit error rate. For example, the process
can be deemed to have converged to an optimal sample time when
further evaluation of the data will lead to no more than a 5%
improvement in the bit error rate.
[0054] Referring now to FIG. 14 there is shown a simplified block
diagram of a radio communication system 1400 that is useful for
understanding the invention. The radio communication system 1400
includes a computer processing unit such as processor 1402. The
system also includes a transceiver 1412 which is comprised of a
transmitter 1414 and a receiver 1416. An antenna 1418 is coupled to
the transceiver to facilitate RF communication. The receiver 1416
is configured to receive RF signals on one or more frequency bands.
The transmitter 1414 is advantageously configured to transmit RF
signals on one or more frequency bands corresponding to a
particular communication protocol. A data communication bus 1402
can be used to communicate digital data as needed among the various
components comprising the radio communication system.
[0055] The communication system can include a user interface such
as display 1404 for communicating information to a user. An input
device 1406 is provided for purposes of allowing a user to enter
control commands and other types of information. The user input
device can include a keypad, a pointing device and any other
suitable types of user input hardware. Suitable communication
interface hardware 1410 is provided to facilitate communication
input and output to the radio communication system. For example,
the communication interface hardware can include a microphone for
detecting user audio (e.g. speech input) and a loudspeaker for
reproducing received audio.
[0056] A memory 1408 is provided for storing programming
instructions which are executed by the processor 1402, and data
needed for operations of the radio communication system 1400. Any
suitable type of memory can be used for this purpose. For example,
the memory 1408 can include one or more of a hard disk drive, a
CD-ROM (compact disk read-only memory), RAM (random-access memory)
or ROM (read-only memory), a flash memory card and so on. Any type
of non-transitory storage medium capable of storing program
instructions and digital data can be used for this purpose. The
programming instructions can include the symbol decoder processing
methods described herein.
[0057] The processor 1402 is comprised of one or more computer
processing elements. For example the computer processing elements
can include a digital signal processor (DSP), a general purpose
microprocessor, a microcontroller, and/or any other processing
device which can be controlled using software or programming
instructions. The operating instructions or computer software
described herein can be stored in in the memory 1408, but can also
reside in memory (not shown) included within the processor 1402.
According to one aspect of the invention, the communication system
1400 can be configured to implement radio communications in
accordance with a P25 communication protocol.
[0058] The transmitter 1414 and receiver 1416 are configured to
facilitate RF communication of data to and to communicate
information bursts or packets from other radio communication
systems. As such, transmitter 1414 and receiver 1416 can include
conventional communication circuitry to enable digital data
transmission over a wireless communication channel. Those skilled
in the art will appreciate that the particular architecture shown
in FIG. 14 is exemplary and merely represents one possible
arrangement of a communication system suitable for implementing the
processing methods described herein. The radio communication system
can be implemented as part of an architecture including hardware
and/or software in accordance with known techniques. Those skilled
in the art will recognize that certain functions of the transmitter
1414 and/or receiver 1416 may be implemented in a processor or
processors, such as the processor 1402.
[0059] The present invention can be realized in one computer
system. Alternatively, the present invention can be realized in
several interconnected computer systems. Any kind of computer
system or other apparatus adapted for carrying out the methods
described herein is suited. A typical combination of hardware and
software can be a general-purpose computer system. The
general-purpose computer system can have a computer program that
can control the computer system such that it carries out the
methods described herein.
[0060] The present invention can take the form of a computer
program product on a computer-usable storage medium (for example, a
hard disk or a CD-ROM). The computer-usable storage medium can have
computer-usable program code embodied in the medium. The term
computer program product, as used herein, refers to a device
comprised of all the features enabling the implementation of the
methods described herein. Computer program, software application,
computer software routine, and/or other variants of these terms, in
the present context, mean any expression, in any language, code, or
notation, of a set of instructions intended to cause a system
having an information processing capability to perform a particular
function either directly or after either or both of the following:
a) conversion to another language, code, or notation; or b)
reproduction in a different material form.
[0061] The methods described herein can be implemented on a
computer system. The computer system can comprise various types of
computing systems and devices, including a server computer, a
client user computer, a personal computer (PC), a tablet PC, a
laptop computer, a desktop computer, a control system, or any other
device capable of executing a set of instructions (sequential or
otherwise) that specifies actions to be taken by that device.
Further, while a single processing device can be used the phrase
"computer system" shall be understood to include any collection of
computing devices that individually or jointly execute a set (or
multiple sets) of instructions to perform any one or more of the
methodologies discussed herein.
[0062] The methods described herein are stored as software programs
in a computer-readable storage medium and are configured for
running on a computer processor. Furthermore, software
implementations can include, but are not limited to, distributed
processing, component/object distributed processing, parallel
processing, virtual machine processing, which can also be
constructed to implement the methods described herein. The term
"computer-readable storage medium" should be taken to include a
single medium or multiple media (e.g., a centralized or distributed
database, and/or associated caches and servers) that store the one
or more sets of instructions. The term "computer-readable storage
medium" shall also be taken to include any medium that is capable
of storing, encoding or carrying a set of instructions for
execution by the machine and that cause the machine to perform any
one or more of the methodologies of the present disclosure.
[0063] The term "computer-readable medium" shall accordingly be
taken to include, but not be limited to, solid-state memories such
as a memory card or other package that houses one or more read-only
(non-volatile) memories, random access memories, or other
re-writable (volatile) memories; magneto-optical or optical mediums
such as a disk or tape. Accordingly, the disclosure is considered
to include any one or more of a computer-readable medium as listed
herein and to include recognized equivalents and successor media,
in which the software implementations herein are stored.
[0064] Although the invention has been illustrated and described
with respect to one or more implementations, equivalent alterations
and modifications will occur to others skilled in the art upon the
reading and understanding of this specification and the annexed
drawings. In addition, while a particular feature of the invention
may have been disclosed with respect to only one of several
implementations, such feature may be combined with one or more
other features of the other implementations as may be desired and
advantageous for any given or particular application. Thus, the
breadth and scope of the present invention should not be limited by
any of the above described embodiments. Rather, the scope of the
invention should be defined in accordance with the following claims
and their equivalents.
[0065] Although the invention has been illustrated and described
with respect to one or more implementations, equivalent alterations
and modifications will occur to others skilled in the art upon the
reading and understanding of this specification and the annexed
drawings. In addition, while a particular feature of the invention
may have been disclosed with respect to only one of several
implementations, such feature may be combined with one or more
other features of the other implementations as may be desired and
advantageous for any given or particular application. Thus, the
breadth and scope of the present invention should not be limited by
any of the above described embodiments. Rather, the scope of the
invention should be defined in accordance with the following claims
and their equivalents.
* * * * *