JP3355147B2 - Automatic frequency control method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a transmission system in which a signal having a frame structure in which a unique word of a known pattern is added for both transmission and reception for each transmission data of a fixed length from the transmission side, and transmitted as a transmission signal. In particular, the present invention relates to a receiving circuit applied to a line in which a multipath with a relatively long delay exists such as an HF band and a relatively large frequency error exists between transmission and reception, such as an HF band. It relates to an automatic frequency control method to be added.

[0002]

2. Description of the Related Art In digital radio communication, a modulated signal is subjected to quadrature quasi-synchronous detection by a fixed oscillator and demodulated by digital signal processing. However, a beat component due to a frequency error occurs in the signal after detection due to a setting error of a local oscillation frequency between the transmitter and the receiver, a temperature variation, and the like. Therefore, it is necessary to compensate for this in order to obtain a correct demodulated signal.
Therefore, in the case of normal digital wireless communication, for example,
A signal known on the transmitting and receiving side, such as a unique word for synchronization acquisition or training of an equalizer, is added to the transmission data.

In a receiver, a method of compensating for a frequency error by detecting a phase error of a pattern between a unique word in a received signal and a unique word known on the receiver side has been conventionally known. However, in the case of wideband modulation, the waveform of the unique word in the received signal is greatly distorted due to the influence of frequency selective fading due to multipath propagation, so that the above method cannot accurately detect the phase error and cannot compensate for the frequency error. . So, usually,
Adaptive equalization processing is applied to prevent frequency-selective fading at the time of wideband transmission. The frequency error that can be compensated by this equalization processing is up to the maximum Doppler frequency that can be compensated by the equalization processing.

In addition to the above-described adaptive equalization processing, other methods for compensating for a frequency error even during wideband transmission include a correlation peak detection type AFC, a tan ^-1 type AFC, and a composite type AFC. Since these are known, these three will be briefly described. FIG. 4 is a block diagram showing a correlation peak detection type AFC circuit of the receiver. On the transmitter side, quadrature modulation is performed with the unique word added, and a signal including the I phase and the Q phase is transmitted. On the receiver side receiving this transmission signal, as shown in FIG. 4, the quadrature detection section 101 performs quadrature detection on the received signal and separates the received signal into an I phase and a Q phase. The complex correlation calculator 104 receives the I-phase signal and the unique word 102 deliberately offset by + fHz, calculates the complex correlation between the two, and the complex correlation calculator 105 calculates the Q-phase signal and
Unique word 104 intentionally offset by -fHz
And calculate the complex correlation between them. The power calculator calculates the sum of squares of the complex correlation as shown below.

That is, the received signal: a _n + jb _n (n = 0,1,2,2
~, P) offset _{signal: c n + jd n (n} = 0,1,2,
~, P)

(Equation 1) Here, (n = 0, 1, 2,..., P).

Here, the method of deriving the unique wad offset by ± f Hz is expressed by the following equations (2) and (3).
Is shown below. First, an offset curve of ± fHz corresponding to the unique word is calculated (eg, the number of unique word symbols = 28 symbols, keying = 1
500 baud). Imaginary part: y = sin (2π · (± f / 1500) · X) Real part: z = cos (2π · (± f / 1500) · X) (2). (X = 0, 1, 2,..., 27) Next, complex multiplication of the unique word and the result calculated by equation (2) is performed as shown in equation (3) (example: amplitude of unique word) a) Unique word: a + ja Offset curve: z + ji (from equation (2)) Complex multiplication: (az-a) + j (az + ay) (3)

[0007] Unique words offset by ± fHz by the above method (eg, ± 40 Hz, ±
50 Hz, ± 60 Hz, etc.) and a complex correlation operation result with a unique word in a signal that intentionally gives an offset to the received signal, the positive / negative complex correlation ratio is calculated according to the following equation. [Hz] offsttRESUL
T) / (-f [Hz] offsttRESULT) where + f [Hz] offsttRESULT: The complex correlation square sum of the unique word offset by + fHz and the input signal intentionally offsetting the received signal. -F [Hz] offsttRESULT : The complex correlation square sum of the unique word offset by −fHz and the input signal intentionally offset from the received signal. Further, referring to FIG. 5, by performing a complex correlation operation with a unique word offset by ± 50 Hz, a frequency error generated in the case of Doppler shift or SSB detuning when the aircraft pass each other in the HF band. Can be compensated for.

Based on the above result, a complex correlation sequence, which is a complex correlation result between the unique word of the received signal and the unique word offset by ± 50 Hz, is output from the complex correlation operation units 104 and 105. Complex correlation operation unit 10
The complex correlation sequences output from 4 and 105 are input to power calculators 106 and 107, and the sum of the squares of the power, which is the power value, is calculated, and input to power ratio calculator 108.
The power ratio calculator 108 calculates the positive / negative complex correlation ratio and outputs it to the frequency error converter 109. The frequency error converter 109 converts the frequency error information into an averaging unit 110
Output to The averaging processing unit 110 performs averaging, and absorbs a large deviation of a calculation result expected to occur suddenly. Equation (4) for averaging is shown below.

[0009]

(Equation 2) The frequency error information detected in this manner is used as a frequency error correction value for controlling the amount of reverse rotation applied to the received signal at the time of orthogonal detection of the next frame. However, in the automatic frequency control system using only this method, the influence of frequency selective fading cannot be completely removed, so that the BER (Bit Erro
r Rate) characteristics are degraded.

Next, the tan ^-1 type AFC will be described. First, the phase error of the received unique word after detection is detected based on the known unique word for both transmission and reception. The following equation (5) is used to detect the unique word phase error.
That is,

(Equation 3) However, phi _s: 1 per symbol error I: I-phase receiving the unique word Q: Q-phase receiving the unique word U _I: I Phase unique word U _Q: Q-phase unique word N: using the unique word length, the unique word 1 This is performed not only between symbols but also over the unique word length, and is averaged to determine an error per symbol (φ _s ). The phase error can be accurately detected even when the received power is reduced by the averaging process.

Next, the forgetting coefficient λ is set to a value greater than or equal to 0 and less than 1, and the phase error φ detected by the above equation (5) is calculated by the following equation (6).

(Equation 4) Averaging processing is performed. This is because the frequency error rarely fluctuates abruptly between frames, and by averaging, the phase of the received signal due to the selective fading is large and the distortion of the equalized output is large. This is to reduce erroneous detection of errors.

Next, the following equation (7)

(Equation 5)

However, when only this method is used, AFC processing must be performed before inputting to the equalizer. Therefore, the operation must be performed using the unique word of the received signal which has not been subjected to the influence of frequency selective fading. However, the frequency error information obtained as the calculation result is completely different from the actual frequency error, so that the BER characteristic is deteriorated as in the case where only the correlation peak detection type AFC is used.

Next, the above-described correlation peak detection type AFC method is first executed, frequency correction is performed once based on the result, and tan ^-1 type AFC is performed using the unique word in the resulting sequence. There is a method in which the result is returned to the quadrature detector and used for frequency correction of the next frame. However, in this method, when the frequency offset is large,
There is a disadvantage that it takes a long time for accurate frequency error correction, and when the Doppler spread of the frequency selective fading is large (more than the Doppler spread at which the equalizer can operate), a phenomenon occurs that the frequency correction cannot follow.

[0015]

In the conventional automatic frequency control system of the digital radio communication described above, the influence of frequency selective fading cannot be completely eliminated, so that the BER characteristic which is the quality of data transmission deteriorates. When the frequency offset is large, it takes a long time to correct the frequency error accurately, and when the Doppler spread of the frequency selective fading is large, the frequency correction cannot follow.

In view of the above problems, the present invention compensates for a frequency error of several tens to several hundreds of Hz (an order of several% of the carrier frequency) which may occur at the time of Doppler shift at the time of passing each other in the HF band or the like or at the time of SSB detuning. An object of the present invention is to provide an AFC system capable of performing the above.

[0017]

In order to solve the above-mentioned problems, the present invention provides a signal having a frame structure in which a unique word of a known pattern is added to both transmission and reception for each transmission data of a certain length from the transmission side. In an automatic frequency control system of a transmission system in which a modulated transmission signal is transmitted,
A quadrature detecting means for receiving the transmission signal on the receiving side and performing quadrature detection using a sine wave; a unique word calculating and storing means for calculating and storing a unique word intentionally given a frequency offset on the receiving side; A maximum value storage means for performing a cross-complex correlation operation between a word and the calculated and stored unique word, detecting a maximum value among the correlation values, and sequentially storing the maximum value, and a maximum value stored in the maximum value storage means. A maximum value detecting means for detecting a maximum value which is the largest value in the value array, and a first frequency error estimating means for estimating a frequency error of the received signal from a frequency offset when the maximum value is obtained. If, when the frequency error estimated by the first frequency error estimating means determines greater or not than the threshold value given in advance, it is determined that the large An estimating means switching means for outputting the frequency error as a frequency error in the frame of the received signal; and a frequency error estimated by the estimating means switching means by the first frequency error estimating means is larger than a predetermined threshold. If it is determined that there is no frequency error, a second error is estimated by performing a complex correlation operation between the unique word in the received signal after detection and the unique word intentionally giving a frequency offset in both positive and negative directions on the receiving side. Giving a rotation in the opposite direction to the frequency error obtained from the received sequence, based on the frequency error estimated by the frequency error estimating means and the first frequency error estimating means or the second frequency error estimating means. A first frequency correction means for reducing a frequency error included in a reception sequence, and a first frequency correction means. Equivalent means for an equivalent delay distortion contained in the output of the third frequency error estimating the final frequency error from the operation of a known unique word in the unique word and the receiving side of the received signal waveform equivalent by the equivalent means Estimating means, adding the frequency error estimated by the first frequency error estimating means or the second frequency error estimating means and the final frequency error estimated by the third frequency error estimating means, and And a second frequency correction means for controlling a sine wave used in the quadrature detector based on the second frequency correction and correcting a frequency error of the next frame.

According to such a configuration, the cross-complex correlation between the unique word in the received signal and the plurality of unique words calculated and stored by the unique word calculation and storage means is calculated by the maximum value storage means. The local maximum value is detected by the local maximum value detecting means, and the frequency error of the received signal is estimated by the first frequency error estimating means from the frequency offset when the local maximum value is obtained. If this value is larger than the threshold, it is output as a frequency error. If this value is not larger than the threshold, the operation shifts to the operation of the second frequency estimating means. In the second frequency estimating means, a complex correlation operation is performed between the unique word in the received signal and the unique word intentionally given a frequency offset in both the positive and negative directions on the receiving side, and a frequency error is estimated. Equivalent processing is added to these frequency errors, and the final frequency error is estimated by the third frequency error estimating means from the calculation with the unique word. The sine wave used in the quadrature detector is controlled by these frequency errors and the final frequency error to correct the frequency error of the next frame.

[0019]

Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing a configuration of an automatic frequency control circuit to which an automatic frequency control system according to the present invention is applied, and FIG. 2 is a block diagram showing a first frequency error estimating block of a frequency error detection unit in FIG. FIG. 3 is a block diagram showing a second frequency error estimating block of the frequency error detecting section of FIG. As shown in FIG. 1, the quadrature detection unit 10 includes a demultiplexer 11 and a π / 2
It comprises a shift calculator 12 and multipliers 13 and 14. The splitter 11 splits the received signal into two waves. The multiplier 13 receives the received signal from the demultiplexer 11 and the sine wave shifted by π / 2 by the π / 2 shift calculator 12, performs multiplication processing, and outputs an I-phase signal. The multiplier 14 receives the received signal from the duplexer 11 and the sine wave, performs a multiplication process, and outputs a Q-phase signal. Multiplier 13,
The I-phase / Q-phase signal output from 14 is delivered to the frequency error detector 20.

The frequency error detecting section 20 includes first and second frequency error estimating blocks 80 and 90. In the first frequency error estimation block 80 shown in FIG.
The complex correlation calculator 81 calculates a cross-complex correlation sequence between the unique word UW in the received signal separated into the I phase / Q phase and the offset UW calculated by the offset UW calculator 82. The maximum value detection unit 83 detects the maximum value in the cross-complex correlation sequence calculated by the complex correlation calculation unit 81, and stores the detected maximum value in the maximum value array memory 84. Thereafter, the offset counter 85 increments the count, and if the result of the increment is within the specified value, the calculation of the offset UW calculation unit 82 is repeatedly executed. In this case, by setting the increment step and the initial value to desired values, flexibility with respect to the environment can be provided. Here, as an example,
The value corresponding to the initial value -75 Hz and the value corresponding to the increment step of 5 Hz were set. Therefore, the maximum value array memory 84 has a range from -75 Hz to 5 Hz.
In the step, the maximum value of the cross complex correlation between +75 Hz is stored.

Next, the maximum value detecting section 86 detects the maximum value in the maximum value array memory 84. The first frequency error estimating section 87 estimates a frequency error corresponding to the local maximum value detected by the local maximum value detecting section 86. Switching section 88
Determines whether the estimated frequency error is greater than a predetermined threshold, and if the estimated frequency error is greater than a predetermined threshold, the estimated frequency error is used as first frequency error information. Output outside block 80. If the frequency error is not larger than the predetermined threshold value, the estimated frequency error is excluded and the operation of the second frequency error estimation block 90 is started.

In the second frequency error estimating block 90 of the frequency error detecting section 20 shown in FIG. 3, a complex correlation calculating section 91 receives an I-phase / Q-phase signal and a unique word 93 offset by + fHz. Then, these cross-complex correlation operations are performed, and the complex correlation operation unit 92 outputs the I-phase / Q-phase signal and the unique word 9 offset by −fHz.
4 to perform these cross-complex correlation calculations. The power calculators 95 and 96 calculate the power of the complex correlation sequences calculated by the complex correlation calculators 91 and 92, respectively. The power ratio calculator 97 calculates a power ratio from the positive and negative power. The second frequency error estimator 98 estimates a frequency error corresponding to the calculated power ratio. The estimated frequency error is
Frequency error estimation block 9 as first frequency error information
Output outside 0.

The first AFC section 30 uses the first frequency error information obtained from the frequency error detecting section 20 to
The frequency correction is performed by giving a reverse rotation to the received signal. The adaptive equalizer 40 uses the unique word UW in the received signal for training, estimates a transmission channel, adaptively calculates tap coefficients, and uses the calculated tap coefficients to generate a unique word UW of the received signal. The signal is output to the third frequency error estimator 50. The third frequency error estimating unit 50 includes an adaptive equalizer 40
Unique word UW of the equalized received signal which is the output of
Tan ^-1 operation with the unique word 60 to obtain second frequency error information. The operation of the third frequency error estimator 50 is the same as that described in the related art. The second AFC section 70 includes an adder 71 and the above-described quadrature detector 10. The adder 71 receives the first frequency error information from the frequency error detector 20 and the third frequency error information.
And the second frequency error information from the frequency error estimating unit 50 is added, and as the frequency error information, the rotation of the sine wave given to the quadrature detector 10 is controlled to correct the frequency.
It is apparent that the same effect can be obtained by using the means for calculating the phase rotation of the unique word UW of the received signal with respect to the known unique word UW by the equation (5) in the frequency error estimating unit.

[0024]

As described in detail above, the automatic frequency control system according to the present invention employs the cross complex correlation between a unique word in a received signal and a plurality of unique words calculated and stored by the unique word calculation and storage means. Is calculated by the maximum value storage means, the maximum value is detected by the maximum value detection means, and the frequency error of the received signal is estimated by the first frequency error estimation means from the frequency offset when the maximum value is obtained. I do. If this value is larger than the threshold value, it is output as a frequency error. If this value is not larger than the threshold value, the operation shifts to the operation of the second frequency estimating means. The second frequency estimating means estimates a frequency error by performing a complex correlation operation between the unique word in the received signal and the unique word intentionally given a frequency offset in both the positive and negative directions on the receiving side. An equivalent process is applied to these frequency errors, and the final frequency error is estimated by the third frequency error estimating means from the calculation with the unique word. The sine wave used in the quadrature detector is controlled by these frequency errors and the final frequency error, and the frequency error of the next frame is corrected. Even if there is a large Doppler shift that occurs when the aircraft passes each other or a frequency error that can occur when detuning in SSB transmission, there is an effect that these corrections can be made.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an automatic frequency control circuit to which an automatic frequency control method according to the present invention is applied.

FIG. 2 is a block diagram showing a first frequency error estimating block of the frequency error detecting section of FIG. 1;

FIG. 3 is a block diagram illustrating a second frequency error estimating block of the frequency error detecting unit in FIG. 1;

FIG. 4 is a block diagram showing a conventional example of a correlation peak detection type automatic frequency control circuit.

FIG. 5 is a diagram showing the operation of the circuit of FIG. 4 when using unique words offset by ± 50 Hz.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Quadrature detector 11 Demultiplexer 12 π / 2 shift calculator 13, 14 Multiplier 20 Frequency error detection unit 30 First AFC unit 40 Adaptive equalizer 50 Third frequency error estimation unit 60 Unique word 70 Second word AFC unit 71 Adder 80 First frequency error estimation block 83 Maximum value detection unit 86 Local maximum value detection unit 87 First frequency error estimation unit 90 Second frequency error estimation block 91, 92 Complex correlation operation unit 97 Power ratio operation Part 98 Second frequency error estimating part

────────────────────────────────────────────────── ─── Continuation of front page (56) References JP-A-11-112592 (JP, A) JP-A-9-162943 (JP, A) JP-A-10-155004 (JP, A) JP-A 8- 331188 (JP, A) JP-A-3-258147 (JP, A) JP-A-10-190758 (JP, A) JP-A-7-297868 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04L 27/00-27/38

Claims (1)

(57) [Claims] 1. For each transmission data of a fixed length from a transmission side,
In an automatic frequency control system of a transmission system in which a transmission signal modulated with a signal having a frame structure to which a unique word of a known pattern is added for transmission and reception is transmitted, a reception side receives the transmission signal and uses a sine wave. a quadrature detection means for quadrature detection, Uny gave deliberately frequency offset at the receiver
A unique word calculation and storage means for calculating and storing the unique word, and performing a cross-complex correlation operation between the unique word of the received signal and the calculated and stored unique word, respectively, detecting a maximum value of the correlation values, and sequentially storing the same. Maximum value storage means, and a maximum value detection means for detecting a maximum value that is the largest value in the maximum value array stored in the maximum value storage means, and from a frequency offset when the maximum value is obtained. First frequency error estimating means for estimating the frequency error of the received signal, and determining whether the frequency error estimated by the first frequency error estimating means is greater than a predetermined threshold value, If it is determined that the large, an estimation unit switching means for outputting the frequency error as a frequency error in a frame of the received signal, estimating means switching means When it is determined that the frequency error estimated by the first frequency error estimating means is not larger than a predetermined threshold, the unique word in the detected received signal and the positive and negative directions in the receiving side are intentionally determined. A second method of estimating a frequency error by performing a complex correlation operation with a unique word having a frequency offset
Frequency error estimating means, and the first frequency error estimating means, or based on the frequency error estimated by the second frequency error estimating means,
A first frequency correction unit for reducing a frequency error included in the received sequence by giving a rotation in a direction opposite to a frequency error obtained from the received sequence, and a delay distortion included in an output of the first frequency correction unit A third frequency error estimating means for estimating a final frequency error from an operation of a unique word of the received signal whose waveform has been equalized by the equalizing means and a unique word known on the receiving side; Or the frequency error estimated by the second frequency error estimating means and the final frequency error estimated by the third frequency error estimating means are added to each other. An automatic frequency control system comprising: a second frequency correction unit that controls a sine wave to be used and corrects a frequency error of a next frame.