Pulse or digital communications – Receivers – Particular pulse demodulator or detector
Reexamination Certificate
1999-05-19
2002-06-11
Vo, Don N. (Department: 2631)
Pulse or digital communications
Receivers
Particular pulse demodulator or detector
C375S347000, C704S262000
Reexamination Certificate
active
06404827
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to method and apparatus for linear predicting, particularly method for linear predicting on received signals in mobile communications and a receiver.
2. Description of the Related Art
In mobile communications, it is a common practice to utilize a least square method as a linear prediction, for correct demodulation of transmission data requires to predict various errors and parameters such as frequency offsets, reference phases, synchronization and radio channels using a limited number of clues such as known signals and provisional judgment signals.
In a conventional least square method, all observed values are equally weighted in the process of linear predicting. This means that both of probable observed values and less probable observed values influenced equally to the result of the linear predicting.
A description of the above conventional least square method is as followed.
FIG. 1
is a graph showing the principle of the conventional least square method. This
FIG. 1
shows a case of applying six observation points which have regular intervals to the predicting.
FIG. 1
shows observed values
101
to
106
, a predicted result
107
, Euclidean distances
108
to
113
, that is, margins of errors between predicted result
107
and each observed values.
Supposing predicted result
107
and Euclidean distances between observed values
101
to
106
and predicted result
107
as margins of errors
108
to
113
, predicted result
107
that will minimize the sum of squares of margins of errors
108
to
113
can be calculated from expressions (1) and (2) below.
Supposing x(i) is an observation point; y(i), an observed value; “N”, the number of observation points; and “Ax(i)+B”, the linear expression to be obtained, “A” and “B” are calculated as follows.
A
=
N
∑
i
=
0
N
-
1
{
X
(
i
)
Y
(
i
)
}
-
∑
i
=
0
N
-
1
X
(
i
)
∑
i
=
0
N
-
1
Y
(
i
)
N
∑
i
=
0
N
-
1
{
X
(
i
)
2
}
-
(
∑
i
=
0
N
-
1
X
(
i
)
)
2
(
1
)
B
=
-
A
∑
i
=
0
N
-
1
X
(
i
)
+
∑
i
=
0
N
-
1
Y
(
i
)
N
(
2
)
Forcing “A” to 0 and only calculating “B”, then, will give an average of y(i).
In
FIG. 1
, the horizontal axis represents observation points and the vertical axis represents observed values at those observation points. “A” represents the gradient of estimation result
107
and “B” represents the intercept.
FIG. 2
is a block diagram showing a configuration of a first embodiment of a receiver which utilizes the conventional least square method as linear predicting. Receiver
200
shown in
FIG. 2
is one of examples of general receivers which utilize the conventional least square method to estimate a series of received signals.
Receiver
200
shown in
FIG. 2
has observation apparatus
202
and least square prediction apparatus
203
. Adder
201
, which is not included in receiver
200
, is described so as to express that received signals are added by disturbances before input into observation apparatus
202
.
In the configuration, a series of signals is input to adder
201
, which is supposed to be such a series of signals that can be expressed by a linear expression.
During signal propagation, disturbances like thermal noises are multiplied into a series of signals in adder
201
. This result is observed by observation apparatus
202
. Since the observation result includes disturbances, the observation result contains a certain margin of error even if the series of signals is one that can be expressed by a linear expression.
Least square prediction apparatus
203
estimates a prediction series by minimizing the square of the margin of error based on the aforementioned expressions (1) and (2). This allows a value close to the series of signals to be obtained as the prediction series even if disturbances exist.
FIG. 3
is a block diagram showing a configuration of a second embodiment of a receiver which utilizes the conventional least square method as linear predicting.
Receiver
300
shown in
FIG. 3
has antenna
301
, frequency offset compensator
302
, demodulator
303
, frequency offset detector
304
and least square prediction apparatus
305
.
In the configuration, suppose a signal received by antenna
301
already contains disturbances. Frequency offset detector
304
detects frequency offsets from received signals.
Least square prediction apparatus
305
calculates the aforementioned expressions (1) and (2) with the detected frequency offsets to obtain probable frequency offsets, and frequency offset compensator
302
uses this prediction result to compensate received signals. By using the prediction result, demodulator
303
obtains demodulated data of better channel quality with its frequency offset compensated.
FIG. 4
is a block diagram showing a configuration of a third embodiment of a receiver which utilizes the conventional least square method as linear predicting. Receiver
400
shown in
FIG. 4
is one of examples of receivers in which the conventional least square method referred to
FIG. 2
is applied to phase estimation essential to detecting synchronization. In detecting synchronization, as explained in
FIG. 3
, even a small margin of error in frequency offset compensation deteriorates the performance, making the reference phase rotate as time goes on. This phase rotation can be expressed in a linear expression using time and phase, so the conventional least square method can be applied to compensate those phase rotation.
Even if frequency offset compensation is completed, to eliminate the influences of constant phase rotation, calculating the averaged phase and compensating the frequency offsets by utilizing the least square method or simply averaging is necessary.
Receiver
400
shown in
FIG. 4
has antenna
401
, phase compensator
402
, demodulator
403
, phase error detector
404
, and least square prediction apparatus
405
.
In the configuration, suppose a signal received by antenna
401
already contains disturbances. Phase compensator
402
compensates the phase shifts of the received signals based on the prediction result obtained from past received signals and demodulator
403
demodulates those result and obtains demodulated data.
Phase error detector
404
detects phase shifts based on the received signals of which phase shifts have been compensated by phase compensator
402
.
Least square prediction apparatus
405
calculates the aforementioned expressions (1) and (2) with the detected phase errors to obtain probable frequency offsets, and phase compensator
302
uses this prediction result to compensate received signals. By using the prediction result, demodulator
403
obtains demodulated data of better channel quality with its phase shifts compensated.
FIG. 5
is a block diagram showing a configuration of a fourth embodiment of a receiver which utilizes the conventional least square method as linear predicting. Receiver
500
shown in
FIG. 5
is one of examples of receivers in which the conventional least square method referred to
FIG. 2
is applied to synchronization shift estimations. Synchronization shifts are originated due to differences in clock oscillation frequencies between transmission and reception. The least square method can be applied to the estimation of synchronization shifts because the relationship between the synchronization and the synchronization shifts can be expressed by a linear expression using time.
Receiver
500
shown in
FIG. 5
has antenna
501
, synchronization timing adjuster
502
, demodulator
503
, synchronization shift detector
504
and least square prediction apparatus
505
.
In the configuration, suppose a signal received by antenna
501
already contains disturbances. Synchronization shift detector
504
detects the synchronization shifts from the received signals between transmission and reception.
Least square prediction apparatus
505
calculates the aforementioned expressions (1) and (2) with the estimated channel quali
Greenblum & Bernstein P.L.C.
Matsushita Electric - Industrial Co., Ltd.
Vo Don N.
LandOfFree
Method and apparatus for linear predicting does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for linear predicting, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for linear predicting will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2938713