JP3377905B2 - Data receiving device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data receiving apparatus used for digital mobile communication or the like, and more particularly to a data receiving apparatus capable of accurately estimating a carrier frequency offset between transmission and reception under various propagation characteristics. It is a thing.

[0002]

2. Description of the Related Art Carrier frequency offset estimating means for a conventional data receiving apparatus (IEICE Transactions B-II p
Fig. 6 shows the configuration of p.884-895 1992/12). In FIG. 6, the carrier frequency offset estimating means of the data receiving apparatus has an input terminal 6-1 to which a received signal is input, and several kinds of frequencies obtained by previously engraving frequencies at which frequency offset is likely to exist at fine intervals δω. The first phase rotation unit 6-2 that rotates so as to compensate the frequency, the correlation calculation unit 6-3 that performs correlation calculation with the rotated signal, and the maximum correlation value are extracted, and the phase rotation amount at that time is calculated. Maximum value detecting section 6-4 for detecting, delay section 6-5 for outputting received signal for each symbol (or 1 bit), and maximum value detecting section 6-4
It is composed of a second phase rotation unit 6-6, which is subjected to phase rotation so as to compensate the phase rotation amount extracted in 1., and an output terminal 6-7 which outputs a reception signal in which the frequency offset is compensated. .

Next, the operation of the above conventional example will be described (see the flowchart of FIG. 20).

In step 1, when the received signal is input to the input terminal 6-1, the first phase rotation unit 6-2 divides the frequency offset range, which is considered to exist in advance, into a certain interval δω. The received signal is rotated to compensate for some of the resulting frequency offsets.

In step 2, the correlation calculator 6-3
Performs a correlation calculation between the received signal and the known signal which are each rotated.

In step 3, the maximum value detector 6-4
Determines that the maximum correlation value is extracted from each correlation value obtained in step 2 and the phase rotation amount at that time is the frequency offset applied at that time to be compensated.

In step 4, the second phase rotation unit 6
-6 multiplies the received signal output from the delay unit 6-5 for each symbol (or 1 bit) by the amount of phase rotation to be compensated obtained in step 3, and outputs it to the output terminal 6-7.

As described above, the carrier frequency offset estimating means of the conventional data receiving apparatus estimates the frequency offset resulting from the frequency difference between transmission and reception by using the correlation calculation process.

[0009]

However, the conventional carrier frequency offset estimating means of the data receiving apparatus cannot cope with various propagation situations, so that it has a problem of having a considerably large estimation error.

Particularly, in a data receiving apparatus having a plurality of branches, if the conventional carrier frequency offset estimating means is applied to each branch, compensation is performed with an offset having a large estimation error in each branch when noise is large. Therefore, there is a problem that the deterioration of the bit error rate due to the estimation error becomes large.

The present invention solves these conventional problems, by multiplying the frequency offset value after estimation by an impulse power component or by using the least squares method from the error component of the demodulated signal. An object of the present invention is to provide a data receiving apparatus which has a carrier frequency offset estimating means which improves the accuracy of frequency offset estimation and is less susceptible to propagation characteristics.

[0012]

In order to achieve the above-mentioned object, the present invention estimates a frequency offset by using a frequency offset amount estimated by a correlation operation and impulse power obtained at that time, and estimates the frequency offset. Is performed after the compensation is performed, and in order to further improve the estimation accuracy, the frequency offset is further estimated from the signal demodulated after compensating the estimated frequency offset by using the least square method, It is possible to realize a data receiving device having a frequency offset estimating means with high accuracy and less likely to be affected by propagation characteristics.

Further, in a data receiving apparatus having a plurality of branches, it is assumed that the reception state in each branch is estimated to be higher as the frequency offset estimated in the branch having a larger impulse power obtained by the correlation calculation has a higher estimation accuracy. Further, it is possible to realize a data receiving apparatus which has a frequency offset estimating means which can estimate a frequency offset with higher accuracy and which is less affected by propagation characteristics with higher accuracy.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT FIG. 1 shows the configuration of a first embodiment of the present invention. In FIG. 1, the carrier frequency offset estimating means of the data receiving apparatus is connected to the phase rotating section 1-2 and the offset compensating sections 1-6-1 to 1-6-n, and the input terminal 1-to which the received signal is input. 1-1 ~
1-1-n, a known signal and a phase rotation unit 1-2 that multiplies a certain frequency offset amount, and a known signal and a reception signal that are connected to the phase rotation unit 1-2 and are multiplied by a certain frequency offset amount. Correlation calculation unit 1-3-1 to 1-3-1
n and connected to the correlation calculation units 1-3-1 to 1-3-3-n,
A frequency offset extraction unit 1-4 that estimates a frequency offset in each branch using the impulse power obtained by each correlation calculation unit, and a frequency offset extraction unit 1-
4, a first estimated value control unit 1-5 for obtaining a frequency offset in the data receiving apparatus using the maximum impulse power and the estimated frequency offset in each branch, and input terminals 1-1-1 to 1-1. −1-n and the first estimation value control unit 1-5, and offset compensation units 1-6-1 to 1-6-n for compensating the received signal for the estimated frequency offset, and the frequency Output terminals 1-7-1 to 1-7 for outputting a signal in which the offset amount is compensated
-N and.

Next, the operation of the first embodiment will be described. First, when a received signal is input to the input terminals 1-1-1 to 1-1-n, the phase rotation unit 1-2 calculates the frequency offset amount obtained by engraving the frequency offset in a certain range in a possible frequency offset range. Multiply known signals.

For example, a frequency offset in the range of ± 1 kHz is considered, and if the frequency offset amount is multiplied by 500 Hz, first, -1 kHz, then -500 Hz,
The offset amount is sequentially applied to the known signal at 0 Hz, 500 Hz, and 1 kHz.

Next, the correlation calculation units 1-3-1 to 1-3-n
Then, the phase rotation unit 1-2 calculates the correlation power between the known signal multiplied by the frequency offset amount and the received signal to calculate the impulse power for each frequency offset, and the frequency offset extraction unit 1-4 calculates the maximum in each branch. The impulse power is detected (the maximum impulse power obtained in each branch is defined as imp1, imp2, ..., impn), and the frequency offset amount applied to the known signal at that time is applied to the frequency offset in that branch ( df1, d
f2, ..., dfn), and the estimated value control unit 1-5
The frequency offset Δf in the data receiving apparatus is estimated by performing the calculation of the following formula 1 on the estimated frequency offset of each branch obtained by the frequency offset extraction unit 1-4. Then, the first offset compensator 1-
In 6-1 to 1-6-n, the frequency offset estimated by the estimated value control unit 1-5 is compensated for the received signal and output to the output terminals 1-7-1 to 1-7-n.

[0018]

[Equation 1] As described above, according to the first embodiment, since all frequency offsets in a possible range are targeted, it is possible to estimate even a considerably large frequency offset, and further, the estimated values in each branch are By contributing the impulse power at that time, the estimated value of the branch that has been received in good condition greatly contributes, and there is an advantage that the frequency offset can be accurately estimated.

FIG. 2 shows the configuration of the second embodiment of the present invention. In FIG. 2, the carrier frequency offset estimating means of the data receiving device is connected to the second offset compensating units 2-2-1 to 2-2-n, and the input terminals 2-1-1 to which the received signal is input. 2-1-n, input terminals 2-1-1 to 2-1-n, and second estimated value control unit 2-7
Frequency compensation is performed on the received signals input from the input terminals 2-1-1 to 2-1-n based on the phase rotation amount estimated by the second estimated value control unit 2-7. The second offset compensating units 2-2-1 to 2-2-n and the demodulating unit 2- that is connected to the second offset compensating units 2-2-1 to 2-2-n and demodulates the received signal 3 and the demodulation unit 2-3
Connected to the demodulator 2-3, the signal demodulated by the demodulator 2-3 is discriminated into 0 and 1 data, and the demodulator 2-3 and the discriminator 2 are connected.
-4, a phase rotation extraction unit 2-5 for obtaining a phase rotation amount from an error between the demodulated signal demodulated by the demodulation unit 2-3 and the signal identified by the identification unit 2-4, and a phase rotation extraction unit 2-
5 is connected to the approximate expression estimating unit 2-6 for obtaining an approximate expression for estimating the amount of phase rotation, and the approximate expression estimating unit 2-6 is connected to the accuracy of the amount of phase rotation estimated from the approximate expression. It is composed of a second estimated value control unit 2-7 for raising and an output terminal 2-8 which is connected to the identifying unit 2-4 and outputs the identified signal.

Next, the operation of the second embodiment will be described. First, when a reception signal is input to the input terminals 2-1-1 to 2-1 -n, the second offset compensating unit 2-2
In -1 to 2-2-n, the frequency offset is compensated for the received signal based on the phase rotation amount estimated by the second estimated value control unit 2-7 (the initial value is 0). The compensated reception signal is demodulated in the demodulation unit 2-3, identified by the identification unit 2-4, and the identified signal is output to the output terminal 2-.
It is output from 8.

On the other hand, in the phase rotation extraction unit 2-5, as shown in FIG. 7, the signal demodulated by the demodulation unit 2-3 is degenerated to (1, 0) and its arctan is taken to obtain the phase rotation amount. And the approximation formula estimation unit 2-6 updates the approximation formula for estimating the phase rotation amount by the least squares method using a plurality of symbols of the estimated phase rotation amount. In order to further improve the estimation accuracy of the phase rotation amount estimated by the estimated value control unit 2-7 using the approximate expression,
If the difference from the estimated value from the previous time is larger than a certain value, it is determined that the estimation error is large, and control such as not considering is added.

As described above, according to the second embodiment, since the frequency offset in each branch is compensated for each symbol, the frequency offset can be accurately compensated in fine increments, and therefore the estimation error is caused. There is an advantage that the data receiving apparatus can be realized with almost no deterioration of the bit error rate. Further, the first embodiment is suitable for estimation in a wide range, for example, when a frequency offset of ± 15 kHz is considered, but rough estimation accuracy (rough estimation accuracy under fading and noise is ±).
On the other hand, the present embodiment has an advantage that the estimation can be performed with high accuracy (the estimation accuracy is approximately within ± 200 Hz under fading or noise) since the estimation is performed while updating the iterative approximation formula. .

FIG. 3 shows the configuration of the third embodiment of the present invention. In FIG. 3, the carrier frequency offset estimating means of the data receiving apparatus is connected to the second offset compensating unit 3-2, and the input terminal 3-1 to which the received signal is input, the input terminal 3-1 and the second terminal. Estimated value control unit 3-
A second offset compensating unit that is connected to 11 and compensates a frequency offset in the received signal input from the input terminal 3-1 based on the phase rotation amount estimated by the second estimated value control unit 3-11. 3-2, a phase rotation unit 3-3 that multiplies a known signal by a certain frequency offset amount, and a correlation between a known signal and a reception signal that are connected to the phase rotation unit 3-3 and are multiplied by a certain frequency offset amount. A correlation calculation unit 3-4, a frequency offset extraction unit 3-5 connected to the correlation calculation unit 3-4, which estimates the frequency offset in each branch using the impulse power obtained by each correlation calculation unit, Two
Offset compensator 3-2 and frequency offset extractor 3
A first offset compensating unit 3-6 connected to -5 to compensate the received signal for the estimated frequency offset.
Connected to the first offset compensating unit 3-6, which demodulates the received signal, and a signal demodulated by the demodulating unit 3-7, which is connected to the demodulating unit 3-7. The identification unit 3-8 for identifying data, the demodulation signal connected to the demodulation unit 3-7 and the identification unit 3-8, demodulated by the demodulation unit 3-7, and the signal identified by the identification unit 3-8. A phase rotation extraction unit 3-9 that obtains the amount of phase rotation from the error, and an approximate expression estimation unit 3-10 that is connected to the phase rotation extraction unit 3-9 and that obtains an approximate expression for estimating the amount of phase rotation. Connected to the estimation unit 3-10,
It is connected to a second estimated value control unit 3-11 for increasing the accuracy of the phase rotation amount estimated by the approximation formula, and an identification unit 3-8,
And an output terminal 3-12 for outputting the identified signal.

Next, the operation of the third embodiment will be described. First, when the received signal is input to the input terminal 3-1, the second offset compensator 3-2 uses the frequency offset based on the phase rotation amount estimated by the second estimated value controller 3-11. Is compensated for the received signal (initial value is 0). Next, for the known signal, the phase rotation unit 3-3
The frequency offset in the possible range is multiplied by the amount of frequency offset obtained by cutting at a certain interval.

For example, a frequency offset in the range of ± 1 kHz is considered, and if the frequency offset amount is multiplied by 500 Hz, first, -1 kHz, then -500 Hz,
The frequency offset amount is sequentially applied to the known signals of 0 Hz, 500 Hz, and 1 kHz. Then, the correlation calculation unit 3-4 calculates the correlation between the known signal multiplied by the frequency offset amount in the phase rotation unit 3-3 and the received signal, calculates the impulse power for each frequency offset, and the frequency offset extraction unit. In 3-5, detect the maximum impulse power,
The frequency offset amount applied to the known signal at that time is used as the estimated frequency offset.

Then, the first offset compensator 3-6
Then, the frequency offset estimated by the frequency offset extraction unit 3-5 is compensated for the received signal. The compensated reception signal is demodulated in the demodulation unit 3-7, and the identification unit 3-
8 and the identified signal is output from the output terminal 3-12.

On the other hand, in the phase rotation extraction unit 3-9, the signal demodulated by the demodulation unit 3-7 as shown in FIG.
The phase rotation amount is estimated by degenerating the phase rotation amount into 0) and taking the arctan, and the approximate expression estimation unit 3-10 uses the least squares method by using a plurality of symbols of the estimated phase rotation amount to calculate the phase rotation amount. Update the approximation for estimating
In order to further improve the estimation accuracy of the phase rotation amount estimated by using the approximate expression in the estimated value control unit 3-11, if the difference from the previous estimated value is larger than a certain value, the estimation error is It is judged to be large and control such as not considering is added.

As described above, according to the third embodiment, since all the frequency offsets in a possible range are targeted, it is possible to estimate even a considerably large frequency offset, and further, it is possible to sequentially estimate every symbol. Since the frequency offset in each branch is compensated for, the frequency offset can be accurately compensated in fine increments, and the advantage that the data receiving apparatus can be realized with almost no deterioration in the bit error rate due to the estimation error of the frequency offset is provided. Have. Therefore, when it is considered that there is a large frequency offset, the first embodiment estimates with a rough estimation error, and the second embodiment takes time to update the approximate expression until the estimation error becomes small. In contrast to this, in the present embodiment, the frequency offset once estimated by the correlation calculation is compensated, and the frequency offset that cannot be completely compensated is estimated by the least square method. Therefore, the estimation can be performed accurately and the estimation error is small. It has the effect that the update of the approximate expression until becomes.

FIG. 4 shows the configuration of the fourth embodiment of the present invention. In FIG. 4, the carrier frequency offset estimating means of the data receiving device is connected to the second offset compensating units 4-2-1 to 4-2-n, and the input terminals 4-1-1 to 4-1-1 to which the received signals are input. 4-1-n, input terminals 4-1-1 to 4-1-n, and second estimated value controller 4-12
The frequency offset is compensated for on the basis of the amount of phase rotation estimated by the second estimated value control unit 4-12 in the received signals input from the input terminals 4-1-1 to 4-1 -n. The second offset compensating units 4-2-1 to 4-2-n, the phase rotation unit 4-3 for multiplying the known signal by a certain frequency offset amount, and the phase rotation unit 4-3 are connected to each other, It is connected to the correlation calculation units 4-4-1 to 4-4-n for correlating the known signal multiplied by the offset amount and the received signal, and the correlation calculation units 4-4-1 to 4-4-n, Frequency offset extraction unit 4 that estimates the frequency offset in each branch using the impulse power obtained in each correlation calculation unit
-5, and a first estimation value control unit 4-6 connected to the frequency offset extraction unit 4-5, which uses the maximum impulse power and the estimated frequency offset in each branch to obtain the frequency offset in the data receiving device. ,
Second offset compensator 4-2-1 to 4-2-n and first offset compensator
Connected to the estimated value control unit 4-6 for compensating the received signal for the estimated frequency offset, and the first offset compensating unit 4-7-1 to 4-7-n. 7-1 to 4-7-n and a demodulation unit 4-8 for demodulating a received signal, and a demodulation unit 4-8 connected to the demodulation unit 4-8.
A discrimination unit 4-9 that discriminates the signal demodulated by -8 into 0 and 1 data, discriminates from the demodulation signal that is connected to the demodulation unit 4-8 and the discrimination unit 4-9, and is demodulated by the demodulation unit 4-8. The phase rotation extraction unit 4-10, which obtains the amount of phase rotation from the error with the signal identified by the unit 4-9, and the phase rotation extraction unit 4-10, which is connected, obtains an approximate expression for estimating the amount of phase rotation. Approximation formula estimation unit 4
-11, connected to the approximate expression estimation unit 4-11, connected to a second estimated value control unit 4-12 for increasing the accuracy of the phase rotation amount estimated by the approximate expression, and an identification unit 4-9, And an output terminal 4-13 for outputting the identified signal.

The operation of the fourth embodiment will be described next. First, when the reception signal is input to the input terminals 4-1-1 to 4-1 -n, the second offset compensating unit 4-2 is provided.
In -1 to 4-2-n, the frequency offset is compensated for the received signal based on the phase rotation amount estimated by the second estimated value control unit 4-12 (initial value is 0). Next, the known signal is multiplied by the frequency offset amount obtained by engraving the frequency offset in a possible range at a certain interval in the phase rotation unit 4-3.

For example, a frequency offset in the range of ± 1 kHz is considered, and if the frequency offset amount is multiplied by 500 Hz, first, -1 kHz, then -500 Hz,
The offset amount is sequentially applied to the known signal at 0 Hz, 500 Hz, and 1 kHz. Then, the correlation calculation units 4-4-1 to 4-4
-N, the phase rotation unit 4-3 performs correlation calculation between the known signal multiplied by the frequency offset amount and the received signal to calculate impulse power for each frequency offset, and the frequency offset extraction unit 4-5 calculates each branch. , The maximum impulse power in each branch is detected (the maximum impulse power in each branch is imp1, imp2, ..., Impn), and the frequency offset amount applied at that time is applied to the frequency offset (df1, df2, ...
, Dfn), and the first estimated value control unit 4-6
The frequency offset in the data receiving apparatus is estimated by performing the calculation of the above-mentioned mathematical expression 1 on the estimated frequency offset of each branch obtained by the frequency offset extraction unit 4-5. Then, the first offset compensator 4-7-
In 1 to 4-7-n, the frequency offset estimated by the first estimated value control unit 4-6 is compensated for the received signal. The compensated reception signal is demodulated in the demodulation unit 4-8,
The signal identified by the identifying unit 4-9 and the identified signal is output terminal 4
Output from -13.

On the other hand, in the phase rotation extraction section 4-10, the signal demodulated by the demodulation section 4-8 as shown in FIG.
The phase rotation amount is estimated by degenerating the phase rotation amount into 0) and taking the arctan, and the approximate expression estimation unit 4-11 uses the least squares method by using a plurality of symbols of the estimated phase rotation amount to calculate the phase rotation amount. Update the approximation for estimating
In order to further improve the estimation accuracy of the phase rotation amount estimated by the estimated value control unit 4-12 using the above approximate expression, the estimation error is large when the difference from the previous estimated value is larger than a certain value. Judgment, and control such as not considering is added.

As described above, according to the fourth embodiment, since all frequency offsets in a possible range are targeted, it is possible to estimate even a considerably large frequency offset, and further, in each branch. By contributing the impulse power at that time to the estimated value, the estimated value of the branch that is in good condition will make a large contribution, and the frequency offset can be accurately estimated.Furthermore, the frequency offset at each branch can be calculated sequentially for each symbol. Since the frequency offset is compensated, the frequency offset can be accurately compensated in fine steps, and the data receiving apparatus can be realized with almost no deterioration of the bit error rate due to the estimation error of the frequency offset.

Therefore, when it is considered that there is a large frequency offset, the first embodiment estimates with a rough estimation error, and the second embodiment updates the approximate expression until the estimation error becomes small. In the third embodiment, since a single branch is used, the frequency offset is once estimated with a rough estimation accuracy only by the impulse power in that branch. Since the estimated frequency offset in the branch with a good state is made to greatly contribute to the estimated value of the data receiving apparatus, it is possible to compensate with the frequency offset estimated with higher accuracy than in the third embodiment, and further compensation is possible. Since the frequency offset which was not present is estimated by the least squares method, the estimation error becomes smaller than that in the second embodiment. Approximate expression for updating has the effect of requiring a shorter time.

The operation of the fifth embodiment will be described with reference to FIG. Here, as an example, a case where the number of branches is 4 will be taken. First, when a signal having a data structure as shown in FIG. 5 is received, a known signal portion (UW portion in FIG. 5) whose data pattern is known in advance has a range within a possible range in the phase rotation portion of the data receiving apparatus. Frequency offset between transmission and reception (-n * δω + ω0 ~ +
n * δω + ω0; for example, ± 1 kHz at a certain step width (δ
ω; for example, 500 Hz) (step 1).

Next, the known signal multiplied by the frequency offset amount in each branch is correlated with the known signal portion of the received signal in each branch (step 2). The frequency offset amount by which the known signal is subjected to the processes of step 1 and step 2 becomes -n * δω + ω0, for example, -1 kHz to + n * δω + ω0, for example, +1 kHz (2 * n +
1) Repeat, for example, 5 times at 500 Hz intervals.

Then, for each branch, each (2 * n +
1) The maximum value is detected among the impulse powers obtained at the time of the correlation calculation (step 3), and the frequency offset amount applied to the known signal having the correlation at that time is set as the estimated frequency offset in the branch (step). 4).

In this way, the frequency offsets estimated for each branch are averaged to obtain the final estimated frequency offset (step 5), and the estimated frequency offset is compensated for the received signal (step 6).

As described above, according to the fifth embodiment, since all frequency offsets in a conceivable range are targeted, it is possible to estimate even a considerably large frequency offset, and further, estimation in each branch. By averaging the values, there is an advantage that the frequency offset can be accurately estimated without bias for each branch.

The operation of the sixth embodiment will be described with reference to FIG. Here, as an example, a case where the number of branches is 4 will be taken. First, when a signal having a data structure as shown in FIG. 5 is received, a known signal portion (UW portion in FIG. 5) whose data pattern is known in advance has a range within a possible range in the phase rotation portion of the data receiving apparatus. Frequency offset between transmission and reception (-n * δω + ω0 ~ +
n * δω + ω0) is multiplied by a certain step size (δω) (step 1).

Next, the known signal multiplied by the frequency offset amount in each branch is correlated with the known signal portion of the received signal in each branch (step 2). The processes of step 1 and step 2 are repeated (2 * n + 1) times in which the frequency offset amount applied to the known signal changes from -n * δω + ω0 to + n * δω + ω0.

Then, for each branch, each (2 * n +
1) The maximum value of the impulse powers obtained at the time of the correlation calculation is detected (step 3), and the frequency offset applied to the known signal having the correlation at that time is set as the estimated frequency offset in the branch (step 4). ).

Thus, the variance of the frequency offset estimated for each branch is obtained (step 5). here,
If the variance obtained in step 5 is greater than or equal to a predetermined threshold value, the final estimated frequency offset is 0, and if the variance is less than a certain threshold value,
The final estimated frequency offset is used as the average value of the estimated frequency offsets obtained in each branch (step 6), and the estimated frequency offset is compensated for the received signal (step 7).

As described above, according to the sixth embodiment, since all frequency offsets in a possible range are targeted, it is possible to estimate even a considerably large frequency offset, and further, the fifth embodiment. Compared to the form in which the estimated value is the average value in each branch, by calculating the variance of the estimated value in each branch, the variation between branches can be taken into consideration, and the frequency offset can be estimated accurately without bias for each branch. It has the advantage of being able to.

The operation of the seventh embodiment will be described with reference to FIG. Here, as an example, a case where the number of branches is 4 will be taken. First, when a signal having a data structure as shown in FIG. 5 is received, a known signal portion (UW portion in FIG. 5) whose data pattern is known in advance,
In the phase rotation unit of the data receiving apparatus, frequency offset (-n * δω + ω0 ~
+ N * δω + ω0) is multiplied for each step size (δω) (step 1).

Next, the known signal multiplied by the frequency offset amount in each branch is correlated with the known signal portion of the received signal in each branch (step 2).

The processes of step 1 and step 2 are
The amount of frequency offset applied to the known signal is -n * δω +
Repeat (2 * n + 1) times from ω0 to + n * δω + ω0. After that, for each branch, the maximum value of the impulse powers obtained at each (2 * n + 1) times of correlation calculation (the maximum value in branch 1 is imp1, the maximum value in branch 2 is imp2, and the maximum value in branch 3 is The maximum value is assumed to be imp3 and the maximum value in branch 4 is assumed to be imp4) (step 3), and the frequency offset amount applied to the known signal correlated at that time is set as the estimated frequency offset in that branch (branch). The estimated value at 1 is df1, the estimated value at branch 2 is df2, the estimated value at branch 3 is df3, and the estimated value at branch 4 is df4) (step 4).

Thus, the formula 1 is calculated using the frequency offset estimated for each branch and the impulse power at that time, and the calculation result is used as the final estimated frequency offset (step 5), and the estimated frequency of the received signal is calculated. Compensate for the offset (step 6).

As described above, according to the seventh embodiment, since all frequency offsets in a possible range are targeted, it is possible to estimate even a considerably large frequency offset, and further, the fifth and fifth embodiments are performed. In the sixth embodiment, the estimation is performed without considering a branch with a good reception state, whereas in the seventh embodiment, the impulse power at that time is contributed to the estimated value in each branch. The estimated value of the branch that has been received in good condition greatly contributes to the advantage that the frequency offset can be accurately estimated.

The operation of the eighth embodiment will be described with reference to FIG. First, the received signal is demodulated (step 1), and as shown in FIG. 7, the demodulated signal is degenerated to (1, 0) and the arctan thereof is calculated to obtain the phase rotation amount (step 2). An approximation formula for estimating the phase rotation amount by the least-squares method is obtained for the phase rotation amount obtained over the number of symbols (step 3), and the approximation formula obtained by the least-squares method is updated using the phase rotation amount obtained sequentially. (Step 4), the frequency offset is estimated from the phase rotation amount obtained from the obtained approximate expression (Step 5).

As described above, according to the eighth embodiment, while the fifth to seventh embodiments estimate the frequency offset with coarse accuracy in one correlation calculation,
In the eighth embodiment, since the approximate expression for sequentially obtaining the phase rotation amount is updated for each certain symbol, there is an advantage that the frequency offset can be accurately estimated in fine steps.

The operation of the ninth embodiment will be described with reference to FIG. First, the received signal is demodulated (step 1), and as shown in FIG. 7, the demodulated signal is degenerated to (1, 0) and the arctan thereof is calculated to obtain the phase rotation amount (step 2). An approximation formula for estimating the phase rotation amount by the least-squares method is obtained for the phase rotation amount obtained over the number of symbols (step 3), and the approximation formula obtained by the least-squares method is updated using the phase rotation amount obtained sequentially. (Step 4), if the absolute value of the phase rotation amount obtained from the obtained approximate expression is equal to or larger than a threshold value,
When it is judged that a large amount of error is included, the compensation is stopped, and when it is smaller than a certain threshold value, the frequency offset is estimated from the phase rotation amount (step 5) and the estimated frequency offset is compensated (step 6). .

As described above, according to the ninth embodiment, the frequency offset is estimated with a coarse accuracy in a single correlation calculation in the fifth to seventh embodiments.
In the ninth embodiment, the approximation formula for sequentially obtaining the phase rotation amount for each symbol is updated, and in the eighth embodiment, the estimated phase rotation amount obtained from the approximation formula is used for compensation. On the other hand, since it is controlled to stop the compensation when it is determined that a large amount of error is included, there is an advantage that the frequency offset can be more accurately estimated and compensated with a fine step.

The operation of the tenth embodiment will be described with reference to FIG. First, the received signal is demodulated (step 1), and as shown in FIG. 7, the demodulated signal is degenerated to (1, 0) and the arctan thereof is calculated to obtain the phase rotation amount (step 2). An approximation formula for estimating the phase rotation amount by the least-squares method is obtained for the phase rotation amount obtained over the number of symbols (step 3), and the approximation formula obtained by the least-squares method is updated using the phase rotation amount obtained sequentially. (Step 4), if the absolute value of the phase rotation amount obtained from the obtained approximate expression is equal to or larger than a threshold value,
It is determined that a large amount of error is included, the estimated frequency offset is set as the previously estimated frequency offset, and when it is smaller than a certain threshold value, the estimated frequency offset is obtained by estimating the phase rotation amount (step 5).

As described above, according to the tenth embodiment, while the fifth to seventh embodiments estimate the frequency offset with coarse accuracy in one correlation calculation, the tenth embodiment is described. In the embodiment of the present invention, the approximate expression for sequentially obtaining the phase rotation amount is updated for each symbol, and
In the embodiment, the estimated phase rotation amount obtained from the approximate expression is used for estimation, and in the ninth embodiment, the estimated value is set to 0 when it is determined that the estimated phase rotation amount includes many errors. On the other hand, if it is determined that the estimated phase rotation amount includes a large amount of error, the frequency offset hardly changes in a short time of a few symbols. It has the advantage that the frequency offset can be estimated.

The operation of the eleventh embodiment will be described with reference to FIG. In a data receiving device having a plurality of branches, first, in each branch, the reception level of a received signal is measured (step 1), then the received signal is demodulated (step 2), and the demodulated signal is received as shown in FIG. Is degenerated to (1, 0) and its arctan is obtained to obtain the phase rotation amount (step 3), and the phase rotation amount obtained over a certain number of symbols is estimated by the least square method. An approximate expression is obtained (step 4), the approximate expression obtained by the least-squares method is updated using the phase rotation amount that is sequentially obtained (step 5), and the measured reception levels in each branch are all less than a certain threshold value. In this case, the compensation is stopped, and if it is above a certain threshold value, the estimated frequency offset is estimated from the phase rotation amount (step 6) and compensated (step 7).

As described above, according to the eleventh embodiment, while the fifth to seventh embodiments estimate the frequency offset with a coarse accuracy at one time of the correlation calculation, the eleventh embodiment does not. In the embodiment of the present invention, the approximate expression for sequentially obtaining the phase rotation amount is updated for each symbol, and
In the embodiment described above, the estimated phase rotation amount obtained from the approximate expression is used for estimation, and in the ninth embodiment, when it is determined that the estimated phase rotation amount includes many errors, the estimated value is set to 0. In the tenth embodiment, when it is determined that the estimated phase rotation amount includes a large amount of error, the estimated value is set as the previous estimated value, but in the present embodiment, the reception level in each branch is Therefore, since the estimated value when it is determined that the reception state is bad is set to 0, there is an advantage that the frequency offset can be accurately estimated in fine increments.

The operation of the twelfth embodiment will be described with reference to FIG. First, in each branch, the reception level of the received signal is measured (step 1), then the received signal is demodulated (step 2), and the demodulated signal is degenerated to (1, 0) as shown in FIG. Then, by obtaining the arctan, the phase rotation amount is obtained (step 3), and the approximate expression for estimating the phase rotation amount by the least square method is obtained for the phase rotation amount obtained over a certain number of symbols (step 4). The approximate expression obtained by the least-squares method is updated using the amount of phase rotation obtained sequentially (step 5), and if the measured reception levels in each branch are all less than a certain threshold value, the estimated frequency offset is set to the previous value. If the estimated frequency offset is equal to or more than a certain threshold value, the estimated frequency offset is estimated from the phase rotation amount (step 6).

As described above, according to the twelfth embodiment, while the fifth to seventh embodiments estimate the frequency offset with coarse accuracy at one time of the correlation calculation, the twelfth embodiment. In the embodiment of the present invention, the approximate expression for sequentially obtaining the phase rotation amount is updated for each symbol, and
In the ninth embodiment, the estimated phase rotation amount obtained from an approximate expression is used for estimation, and in the ninth embodiment, when it is determined that the estimated phase rotation amount includes many errors, the estimated value is set to 0, and In the tenth embodiment, when it is determined that the estimated phase rotation amount includes many errors, the estimated value is set as the previous estimated value,
In the eleventh embodiment, the reception state is determined from the reception level in each branch, and the estimated value is set to 0 when it is determined that the reception state is bad, whereas in the present embodiment, the frequency offset is several. Since it hardly changes in a short time called a symbol, when it is judged that the reception state is bad,
By using the estimated phase change amount of the previous time, there is an advantage that the frequency offset can be accurately estimated in fine increments.

The operation of the thirteenth embodiment will be described with reference to FIGS. 16 and 17. First, when a signal having a data structure as shown in FIG. 5 is received, a known signal portion (UW portion in FIG. 5) whose data pattern is known in advance has a range within a possible range in the phase rotation portion of the data receiving apparatus. Frequency offset between transmission and reception (-n *
Multiply δω + ω0 to + n * δω + ω0 for each step size (δω) (step 1).

Next, the known signal multiplied by the frequency offset amount in each branch is correlated with the known signal portion of the received signal in each branch (step 2). The processes of step 1 and step 2 are repeated (2 * n + 1) times in which the frequency offset amount applied to the known signal changes from -n * δω + ω0 to + n * δω + ω0.

After that, the maximum value of the impulse powers obtained at the time of each (2 * n + 1) times of correlation calculation is detected (step 3), and the frequency offset amount applied to the known signal at that time is estimated. It is a frequency offset (step 4). Then, the estimated frequency offset of the received signal is compensated (step 5).

Next, the received signal compensated in step 5 is demodulated (step 6), and as shown in FIG. 7, the demodulated signal is degenerated to (1, 0) to obtain its arctan. The phase rotation amount is obtained (step 7), the phase rotation amount obtained over a certain number of symbols is estimated by the least square method, and an approximate expression for estimating the phase rotation amount is obtained (step 8). The approximation formula obtained by the least squares method is updated (step 9), and when the absolute value of the phase rotation amount obtained from the obtained approximation formula is more than a certain threshold value, it is judged that a lot of errors are included. , Compensation is stopped, and when it is smaller than a certain threshold value, the frequency offset is estimated from the phase rotation amount (step 1
0), the estimated frequency offset is compensated (step 11).

As described above, according to the thirteenth embodiment, the fifth to seventh embodiments estimate the frequency offset with coarse accuracy in one correlation calculation, and the eighth to twelfth embodiments. Is the minimum phase rotation amount of the demodulated signal is 2
Compared with the case where the frequency offset is estimated by the multiplication method, in the thirteenth embodiment, since all the frequency offsets in the conceivable range are targeted, it is possible to estimate even a considerably large frequency offset. Since the offset in each branch is compensated for each symbol, there is an advantage that the frequency offset can be accurately compensated in fine increments, and the bit error rate can be realized with almost no deterioration of the bit error rate due to the estimation error of the frequency offset.

The operation of the fourteenth embodiment will be described with reference to FIGS. 18 and 19. Here, as an example, a case where the number of branches is 4 will be taken. First, when a signal having a data structure as shown in FIG. 5 is received, a known signal portion whose data pattern is known in advance (see FIG.
UW) of the data receiving device,
Frequency offset between possible transmission and reception (-n
* Δω + ω0 to + n * δω + ω0) with a certain step width (δω)
Multiply each time (step 1).

Next, the known signal multiplied by the frequency offset amount in each branch is correlated with the known signal portion of the received signal in each branch (step 2). The processes of step 1 and step 2 are repeated (2 * n + 1) times in which the frequency offset amount applied to the known signal changes from -n * δω + ω0 to + n * δω + ω0.

Thereafter, for each branch, each (2 * n +
1) The maximum value of the impulse powers obtained in the correlation calculation (imp1 is the maximum value in branch 1, the maximum value in branch 2 is imp2, the maximum value in branch 3 is imp3, the maximum value in branch 4) Is defined as imp4) (step 3), and the frequency offset amount applied to the known signal obtained by correlation at that time is set as the estimated frequency offset in that branch (estimated value in branch 1 is df1, branch 2 is Df2, the estimated value in branch 3 is df3, and the estimated value in branch 4 is df4) (step 4).

Thus, the equation 1 is calculated using the frequency offset estimated for each branch and the impulse power at that time, and the calculation result is used as the final estimated frequency offset (step 5). Compensate for the offset (step 6).

Next, the received signal compensated in step 6 is demodulated (step 7), and the demodulated signal is degenerated to (1, 0) as shown in FIG. 7 to obtain its arctan. The amount of phase rotation is calculated (step 8), and an approximate expression for estimating the amount of phase rotation by the least squares method is calculated for the amount of phase rotation obtained over a certain number of symbols (step 9). The approximation formula obtained by the least-squares method is updated (step 10), and if the absolute value of the phase rotation amount obtained from the approximation formula obtained is above a certain threshold value, it is judged that there are many errors. , Compensation is stopped, and when it is smaller than a certain threshold value, the frequency offset is estimated from the phase rotation amount (step 1
1) Compensate the estimated frequency offset (step 12).

As described above, according to the fourteenth embodiment, the fifth to seventh embodiments estimate the frequency offset with coarse accuracy at one time of the correlation calculation, and the eighth to twelfth embodiments are performed. Is the minimum phase rotation amount of the demodulated signal is 2
In the fourteenth embodiment, the frequency offset is estimated by a multiplication method, and the frequency offset is estimated with rough estimation accuracy only by the impulse power in the branch in the thirteenth embodiment. Since all possible frequency offsets are targeted, it is possible to estimate even a considerably large frequency offset. Furthermore, by contributing the impulse power at that time to the estimated value in each branch, the branch that is in good condition is received. The estimated value of will contribute greatly and the frequency offset can be estimated with high accuracy.Furthermore, the offset in each branch will be compensated for every symbol sequentially, so the frequency offset can be accurately compensated with a fine step, and the frequency offset Bit error rate due to the estimation error of It has the advantage that it can be realized without.

[0071]

As is clear from the above description of the embodiment, the present invention covers all frequency offsets within a conceivable range, so that it is possible to estimate even a considerably large frequency offset. In addition, the frequency offset can be estimated with high accuracy while sequentially updating the approximate expression for obtaining the phase rotation amount using the least square method, and the impulse power at that time can be contributed to the estimated value in each branch to The frequency offset can be accurately estimated by making a large contribution to the estimated value of the branch that is well received, and furthermore, the offset in each branch is successively compensated for every certain symbol, so it is estimated with high accuracy, By compensating, it is possible to realize a data receiving device that hardly causes deterioration of the bit error rate due to the estimation error of the frequency offset. With a cormorant effect.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a first embodiment of a data receiving device of the present invention,

FIG. 2 is a block diagram showing a configuration of a second embodiment of a data receiving device of the present invention,

FIG. 3 is a block diagram showing a configuration of a third embodiment of a data receiving device of the present invention,

FIG. 4 is a block diagram showing a configuration of a fourth embodiment of a data receiving device of the present invention,

FIG. 5 is a data configuration example used in the embodiment of the present invention,

FIG. 6 is a block diagram showing a configuration of a carrier frequency offset estimating means in a conventional data receiving device,

FIG. 7 is an operation explanatory diagram used in the description of the embodiment of the present invention;

FIG. 8 is a flowchart illustrating frequency offset estimation operation according to the fifth embodiment of the present invention;

FIG. 9 is a flowchart illustrating a frequency offset estimating operation according to the sixth embodiment of the present invention;

FIG. 10 is a flowchart illustrating frequency offset estimation operation according to the seventh embodiment of the present invention;

FIG. 11 is a flowchart for explaining frequency offset estimation operation according to the eighth embodiment of the present invention;

FIG. 12 is a flowchart for explaining frequency offset estimation operation according to the ninth embodiment of the present invention;

FIG. 13 is a flowchart illustrating frequency offset estimation operation according to the tenth embodiment of the present invention;

FIG. 14 is a flowchart illustrating frequency offset estimation operation according to the eleventh embodiment of the present invention;

FIG. 15 is a flowchart illustrating frequency offset estimation operation according to the twelfth embodiment of the present invention;

FIG. 16 is a flowchart illustrating frequency offset estimation operation according to the thirteenth embodiment of the present invention;

FIG. 17 is a flowchart for explaining frequency offset estimation operation according to the thirteenth embodiment of the present invention;

FIG. 18 is a flowchart illustrating frequency offset estimation operation according to the fourteenth embodiment of the present invention;

FIG. 19 is a flowchart illustrating frequency offset estimation operation according to the fourteenth embodiment of the present invention;

FIG. 20 is a flowchart illustrating a conventional frequency offset estimating operation.

[Explanation of symbols]

1-1-1 to 1-1-n, 2-1-1 to 2-1-n input terminal 3-1, 4-1-1 to 4-1-n input terminal 1-2-1 to 1- 2-n, 3-3, 3-9 Phase rotation unit 4-3-1 to 4-3-n, 4-10 Phase rotation unit 1-3-1 to 1-3-n, 3-4 Correlation calculation unit 4-4-1 to 4-4-n, 6-3 Correlation calculation units 1-4, 3-5, 4-5 Frequency offset extraction units 1-5, 4-6 Estimated value control unit 1 (first estimation Value control unit) 1-6-1 to 1-6-n, 3-6 Offset compensation unit 1 4-7-1 to 4-7-n Offset compensation unit 1 (first offset compensation unit) 1-7- 1 to 1-7-n, 2-8 output terminals 3-12, 4-13, 6-7 output terminals 2-2-1 to 2-2-n, 3-2 offset compensation unit 2 4-2-1 ~ 4-2-n Offset compensator 2 (second offset compensator) 2-3 , 3-7, 4-8 Demodulation section 2-4, 3-8, 4-9 Identification section 2-5 Phase rotation extraction section 2-6, 3-10, 4-11 Approximate expression estimation section 2-7, 3 -11, 4-12 Estimated value control unit 2 (Second estimated value control unit) 6-2 Phase rotation unit 1 (First phase rotation unit) 6-4 Maximum value detection unit 6-5 Delay unit 6-6 Phase rotation unit 2 (second phase rotation unit)

─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-8-223240 (JP, A) JP-A-7-297870 (JP, A) JP-A-7-66842 (JP, A) JP-A-7- 183925 (JP, A) JP 7-115445 (JP, A) JP 6-97972 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04L 27/22

Claims (14)

(57) [Claims] 1. In a data receiving device having a plurality of branches, a phase rotation unit for multiplying a known signal by a frequency offset amount obtained by cutting a frequency offset in a conceivable range at a certain interval, and rotation by the phase rotation unit. Correlation calculation unit that takes the correlation between the known signal and the received signal, obtains the maximum impulse power of the impulse power obtained by the correlation calculation unit, the frequency offset applied to the known signal when the maximum impulse power is obtained A frequency offset extraction unit that makes the amount an estimated frequency offset in each branch at that time, an estimated value control unit that finally obtains a frequency offset in the receiving device by using the estimated frequency offset and impulse power in each branch, and a final Off to compensate for the dynamically estimated frequency offset A data receiving device, comprising: a set compensating unit. 2. A demodulation section for demodulating a received signal, an identification section for identifying the demodulated signal, a phase rotation extraction section for extracting a phase rotation amount from an error between the demodulated signal and the identified signal, and a certain number of symbols. The phase rotation amount obtained over
Obtaining an approximate expression that estimates the phase rotation amount with high accuracy by the multiplicative method, further, an approximate expression estimating unit that updates the approximate expression using the phase rotation amount that is sequentially obtained, and the phase rotation amount estimated from the approximate expression. A data receiving device, comprising: an estimated value control unit for increasing estimation accuracy; and an offset compensating unit for estimating and compensating a frequency offset from the finally estimated phase rotation amount. 3. A data receiving device having a single branch, wherein a phase rotation unit for multiplying a known signal by a frequency offset amount obtained by cutting a frequency offset in a conceivable range at a certain interval, and rotation by the phase rotation unit. Correlation calculation unit that takes the correlation between the known signal and the received signal, obtains the maximum impulse power of the impulse power obtained by the correlation calculation unit, the frequency offset applied to the known signal when the maximum impulse power is obtained A frequency offset extraction unit that sets the amount as an estimated frequency offset at that time, a first offset compensation unit that compensates the estimated frequency offset, a demodulation unit that demodulates the received signal in which the estimated frequency offset is compensated, and a demodulated signal The phase rotation amount is determined from the error between the identification unit that identifies the And a phase rotation extraction unit for extracting the phase rotation amount obtained over a certain number of symbols, and an approximate expression for estimating the phase rotation amount with high accuracy by the least square method is obtained, and the phase rotation amount obtained sequentially is used. An approximate expression estimation unit that updates the approximate expression by a second estimation value, and a second estimated value control unit that further increases the estimation accuracy of the phase rotation amount by the approximate expression.
A second offset compensator that estimates and compensates for a frequency offset from the estimated amount of phase rotation, the data receiving apparatus. 4. In a data receiving device having a plurality of branches, a phase rotation unit for multiplying a known signal by a frequency offset amount obtained by cutting a frequency offset in a conceivable range at a certain interval, and rotation by the phase rotation unit. Correlation calculation unit that takes the correlation between the known signal and the received signal, obtains the maximum impulse power of the impulse power obtained by the correlation calculation unit, the frequency offset applied to the known signal when the maximum impulse power is obtained A frequency offset extraction unit that sets the amount as an estimated frequency offset in each branch at that time, and a first estimated value control unit that obtains a frequency offset in the data receiving device by using the estimated frequency offset and impulse power in each branch, A first offset that compensates for the estimated frequency offset. A set compensation unit, a demodulation unit that demodulates the received signal in which the estimated frequency offset is compensated, an identification unit that identifies the demodulated signal, and a phase rotation that extracts the phase rotation amount from the error between the demodulated signal and the identified signal. An approximation formula for estimating the phase rotation amount with high accuracy by the extraction unit and the phase rotation amount obtained over a certain number of symbols is obtained by the least square method, and the approximation formula is updated using the phase rotation amount obtained successively. A second estimated value control unit for further increasing the estimation accuracy of the phase rotation amount by the approximation formula, and a second estimated value offset unit for estimating and compensating the frequency offset from the estimated phase rotation amount. An offset compensation unit, and a data receiving device. 5. In a data receiving device having a plurality of branches, a means for multiplying a known signal by a frequency offset amount in a possible range, a means for obtaining a correlation between a known signal and a received signal multiplied by the frequency offset amount, and the correlation. A means for obtaining the maximum impulse power among the impulse powers obtained by the means for obtaining, and an estimation at that time of the frequency offset amount applied to the known signal when the maximum impulse power was obtained by the means for obtaining the maximum impulse power. Means for setting a frequency offset, means for averaging the estimated frequency offsets obtained in each branch, and finally making the value a frequency offset in the data receiving device, and means for compensating the frequency offset And a data receiving device comprising: . 6. In a receiver having a plurality of branches, means for multiplying a known signal by a frequency offset amount in a possible range, means for obtaining a correlation between a known signal and a received signal multiplied by the frequency offset amount, and the correlation A means for obtaining the maximum impulse power among the impulse powers obtained by the obtaining means, and an estimation at that time of the frequency offset amount applied to the known signal when the maximum impulse power is obtained by the means for obtaining the maximum impulse power. Frequency offset means, means for obtaining the variance of the estimated frequency offset obtained in each branch, and in the data receiving device if the variance obtained by the variance obtaining means is above a certain threshold If the frequency offset is estimated to be 0 and is below a certain threshold, In the case of a means for averaging the estimated frequency offset obtained in each branch as a frequency offset in the data receiving device,
Means for compensating for the frequency offset, and a data receiving device. 7. In a receiver having a plurality of branches, means for multiplying a known signal by a frequency offset amount in a possible range, means for obtaining a correlation between a known signal and a received signal multiplied by the frequency offset amount, and the correlation A means for obtaining the maximum impulse power among the impulse powers obtained by the obtaining means, and an estimation at that time of the frequency offset amount applied to the known signal when the maximum impulse power is obtained by the means for obtaining the maximum impulse power. A means to be a frequency offset amount, the estimated frequency offset amount obtained in each branch, the impulse power of each branch divided by the sum of the impulse power of all branches, multiplied by the value obtained by adding Means for estimating the frequency offset in the data receiving device, and And a means for compensating for the frequency offset, the data receiving apparatus. 8. A means for demodulating a received signal, an identifying means for identifying the demodulated signal, a means for obtaining a phase rotation amount from an error between the demodulated signal and the identified signal, and the phase obtained for a certain number of symbols. A means for obtaining an approximation formula for estimating the phase rotation amount with high accuracy by the least squares method of the rotation amount, and a means for updating the approximation formula obtained by the least squares method using the phase rotation amounts obtained successively are estimated. And means for estimating the frequency offset amount from the phase rotation amount,
A data receiving device comprising: 9. A means for demodulating a received signal, an identifying means for identifying the demodulated signal, a means for obtaining an amount of phase rotation from an error between the demodulated signal and the identified signal, and a means for obtaining the phase rotation amount obtained over a certain number of symbols. A means for obtaining an approximate expression for estimating the phase rotation amount with high accuracy by the least squares method, and a means for updating the approximate expression obtained by the least squares method using the phase rotation amounts sequentially obtained; Compensation is stopped when the absolute value of the phase rotation amount is greater than or equal to a threshold value, and means for estimating a frequency offset amount from the estimated phase rotation amount when the absolute value is less than a certain threshold value, A means for compensating the frequency offset amount, and a data receiving device. 10. A means for demodulating a received signal, an identifying means for identifying the demodulated signal, a means for obtaining a phase rotation amount from an error between the demodulated signal and the identified signal, and a means for obtaining the phase rotation amount over a certain number of symbols. Means for obtaining an approximate expression for estimating the phase rotation amount by the least squares method, and means for updating the approximate expression obtained by the least squares method using the phase rotation amounts sequentially obtained; If the absolute value of the phase rotation amount is more than a certain value, the frequency offset amount used last time is set as the frequency offset amount to be compensated this time, and if it is less than a certain value, the frequency offset amount is calculated from the phase rotation amount estimated this time. And a means for estimating the data receiving apparatus. 11. In a receiver having a plurality of branches, a means for measuring a received signal level, a means for demodulating the received signal, an identifying means for identifying the demodulated signal, and an error between the signal identified as the demodulated signal A means for obtaining the amount of phase rotation, a means for obtaining an approximate expression for highly accurately estimating the amount of phase rotation by the least-squares method for the amount of phase rotation obtained over a certain number of symbols, and the amount of phase rotation that is sequentially obtained A means for updating the approximate expression obtained by the least squares method, and when the received signal level at that time is less than a certain value in each branch, the compensation is stopped, and at least one value in each branch And a means for estimating the frequency offset amount from the estimated phase rotation amount. 12. A receiving device having a plurality of branches, comprising: a unit that demodulates a received signal; an identifying unit that identifies the demodulated signal; and a unit that obtains a phase rotation amount from an error between the demodulated signal and the identified signal. A means for obtaining an approximate expression for estimating the phase rotation amount with high accuracy by the least square method for the phase rotation amount obtained over the number of symbols; and an approximation obtained by the least square method using the phase rotation amount obtained successively. If both the means for updating the formula and the received signal level at that time are less than a certain value in each branch, the frequency offset amount used last time is set as the frequency offset amount to be compensated this time, and it is also one of the respective branches. And a means for estimating a frequency offset amount from the phase rotation amount estimated this time when the value is equal to or more than a value. . 13. In a receiver having a single branch, first, means for multiplying a known signal by a frequency offset amount in a possible range, means for obtaining a correlation between a known signal and a received signal multiplied by the frequency offset amount, A means for obtaining the maximum impulse power among the impulse powers obtained by the means for obtaining the correlation, and a frequency offset amount applied to a known signal when the maximum impulse power is obtained by the means for obtaining the maximum impulse power. Means for estimating the frequency offset, means for compensating the frequency offset, means for demodulating the received signal with the frequency offset compensated, identifying means for identifying the demodulated signal, and error between the signal identified as the demodulated signal To obtain the amount of phase rotation from And a means for obtaining an approximate expression for estimating the phase rotation amount by the least squares method, and a means for updating the approximate expression obtained by the least squares method using the sequentially obtained phase rotation amount. When the absolute value of the phase rotation amount is more than a certain value, the compensation is stopped, and when it is less than a certain value, a means for estimating the frequency offset amount from the estimated phase rotation amount, and the frequency offset amount are compensated. And a means for performing the same. 14. In a receiver having a plurality of branches, first, a means for multiplying a known signal by a frequency offset amount in a possible range, a means for obtaining a correlation between a known signal multiplied by the frequency offset amount and a received signal, A means for obtaining the maximum impulse power among the impulse powers obtained by the means for obtaining the correlation, and a frequency offset amount applied to a known signal when the maximum impulse power is obtained by the means for obtaining the maximum impulse power. Means for setting the estimated frequency offset amount,
The estimated frequency offset amount obtained in each branch, the impulse power of each branch divided by the sum of the impulse power of all branches, multiplied by, the value obtained by adding and estimated as the frequency offset in the data receiving apparatus Means, means for compensating for the frequency offset thereof, means for demodulating the received signal for which the frequency offset has been compensated, identifying means for identifying the demodulated signal,
Means for obtaining the amount of phase rotation from the error between the demodulated signal and the identified signal; means for obtaining an approximate expression for estimating the amount of phase rotation by the least square method for the amount of phase rotation obtained over a certain number of symbols; A means for updating an approximate expression obtained by the least-squares method using the phase rotation amount that is sequentially obtained, and stopping compensation when the absolute value of the estimated phase rotation amount is a certain value or more, and less than a certain value. In some cases, the data receiving apparatus is provided with means for estimating a frequency offset amount from the estimated phase rotation amount, and means for compensating the frequency offset amount.