KR20010064275A - Compensation for the Doppler Frequency Shift using FFT - Google Patents

Abstract

PURPOSE: A Doppler frequency transition compensating method using an FFT(Fast Fourier Transform) is provided to compare signal component powers of two frequencies adjacent to an IF(Intermediate Frequency), and to control an output frequency of a local oscillator. CONSTITUTION: After performing a frequency transition to a base band, a system detects a Doppler frequency transition. According to a detected result, the system compensates a remaining carrier wave frequency of a base band receiving signal. The system performs an FFT(Fast Fourier Transform) for the signal, and calculates a power distribution of a discrete frequency component. The system estimates a remaining carrier wave frequency error. The system reduces noises by using an LPF(Low Pass Filer). The system controls a frequency of a local oscillator, and compensates the remaining carrier wave frequency error. If errors are generated while compensating the remaining carrier wave frequency error, the system returns to a remaining carrier wave frequency compensation step or an error estimating step according to an error kind.

Description

Compensation for Doppler Frequency Shift Using Fast Fourier Transform {Compensation for the Doppler Frequency Shift using FFT}

The present invention relates to a digital wireless communication system in a channel environment in which a Doppler frequency shift exists. More specifically, the present invention relates to a Doppler frequency shift compensation method using a fast Fourier transform.

To compensate for the Doppler frequency shift, the carrier frequency error must first be detected, and then the locally generated carrier frequency must be adjusted based on the detection result. Conventional techniques for detecting the carrier frequency error use a method of comparing the average phase of the received signal during two consecutive sections in the time domain, or a method of comparing the power of two different frequency signal components. The method of comparing the average phase in the time domain should perform a calculation or a complex multiplication operation for calculating a phase value with respect to the complex signal. To compare the power of two different frequency signal components, the frequency components are extracted by using a mixer and two filters or by using a large fast Fourier transformer.

The conventional carrier frequency shift compensation method has a large complexity in implementing a circuit. In addition, when using an interpolator in adjusting the precision of carrier frequency error detection, there is a problem that the complexity is further increased.

Accordingly, the present invention has been made to solve the problems of the prior art as described above, the Doppler frequency using a fast Fourier transform, the complexity is minimized by using a power balancing technique in the frequency domain using a small fast Fourier transformer It is to provide a transition compensation method and apparatus.

Another object of the present invention is to provide a Doppler frequency shift compensation method and apparatus using a fast Fourier transform that can perform fast initial frequency acquisition and precise final frequency compensation by adjusting a correlation operation length in a step before a fast Fourier transform operation.

1 is a flowchart illustrating Doppler transition and PN code timing recovery after receiving a radio frequency signal at a receiver of a wireless communication system using a direct sequence spread spectrum signal (DS-CDMA) scheme;

2 is a block diagram illustrating a receiver of a wireless communication system corresponding to the flowchart of FIG. 1;

3 is a detailed block diagram illustrating a detailed Doppler transition compensation unit;

4 is a graph showing signal component power of a correlator output signal according to a correlator length for various residual carrier frequency errors Δf;

5 is a graph showing the interference state between the elements of the fast Fourier transform result vector with respect to the frequency of a single tone input;

6 is an internal configuration diagram of a fast Fourier transform processor of size 4,

FIG. 7 is a graph illustrating a fast Fourier transform processor and a residual carrier frequency error estimation result characteristic having a simple structure as shown in FIG.

8 is a block diagram illustrating an interlocking structure of PN (Pseudo-random Noise) pseudo timing acquisition by fast Fourier transform-based two-dimensional retrieval and an approximate Doppler transition estimation and a detailed Doppler transition compensation loop;

9 is a flowchart illustrating a fast Fourier transform-based PN code synchronization acquisition and an approximate Doppler transition detector according to an embodiment of the present invention;

FIG. 10 is a flowchart illustrating an interlocking operation between a fast Fourier transform-based detailed Doppler transition compensation unit, a PN code synchronization acquisition unit, and an approximate Doppler transition detection unit according to an embodiment of the present invention.

In order to achieve the above object, the Doppler frequency shift compensation method using the fast Fourier transform according to the present invention, after the frequency shift to the baseband with respect to the received signal, detects the approximate Doppler frequency shift, and the baseband according to the detection result A first step of compensating for the residual carrier frequency of the received signal; Calculating a power distribution of discrete frequency components by performing a fast Fourier transform operation on the approximate Doppler frequency shift detection and compensation signal; Estimating a residual carrier frequency error from the fast Fourier transform calculation result; A fourth step of reducing noise using a low pass filter for the residual carrier frequency error estimation result; A fifth step of compensating for a residual carrier frequency error by adjusting a frequency of a local oscillator according to the low pass filtered result and proceeding to the second step; And a sixth step of repeating from the first or second step according to the type of error when an error occurs when compensating for the residual carrier frequency error.

Preferably, in the first step, if the received signal is a direct sequence spread spectrum signal, frequency shifting is performed to the baseband, and then inversely converted to the received signal by correlating with a locally generated PN code to approximate Doppler frequency shifting. It is characterized in that for detecting.

Preferably, the second step includes a first substep of deliberately generating a correlation between adjacent frequency bins by padding the same number of zeros with the output signal of the first step; And a second sub-step of performing a Fourier transform operation, wherein the third step is performed by using the symmetry of the correlation between adjacent frequency bins of the fast Fourier transform output vectors. The residual carrier frequency error is estimated from the signal component difference.

More preferably, the second sub-step of the second step is characterized by using a fast Fourier transform calculator of size 4 consisting of two adders and two subtractors.

More preferably, in the third step, by applying C [1] and C [2P-1] of the fast Fourier transform operation vector to the following equation to obtain the difference between the magnitude squared values, the residual carrier frequency error is determined. It is characterized by estimating.

[Equation]

More preferably, the third step is to estimate the residual carrier frequency error by applying the C [1] and C [2P-1] of the fast Fourier transform operation vector to the following equation to obtain the magnitude difference. It features.

[Equation]

Preferably, the fourth step comprises: a first sub-step of passing the approximate Doppler frequency shift compensated signal through a wide bandwidth loop filter to quickly converge residual carrier frequency error within a predetermined range within a short time; And if the residual carrier frequency error converges within a predetermined range, narrowing the loop filter bandwidth so as to converge the residual carrier frequency error in a steady state within a target tolerance finally.

More preferably, the fourth step is characterized in that the second sub-step is performed when it is detected that the residual carrier frequency error converges within a predetermined range.

More preferably, in the fourth step, the second sub-step may be performed after forcibly resetting the bandwidth of the loop filter after a time estimated that the residual carrier frequency error has converged within a predetermined range. do.

More preferably, the sixth step may include a first sub-step of generating a residual carrier frequency reconstruction error during compensation of the residual carrier frequency error, and generating an error in PN code timing and Doppler frequency shift compensation for the occurrence error. A second sub-step of checking whether or not, and a PN code timing acquisition step, an approximate Doppler frequency shift detection step, and a residual carrier frequency error estimation step according to the identified occurrence type of error, and then thereafter And a third substep of repeating these steps.

Even more preferably, the first sub-step of the sixth step may include C [1], C [2P-1], and C [0] of the fast Fourier transform result vectors used to detect the residual carrier frequency error. Detecting using a pattern of complex magnitude values.

Even more preferably, in the second sub-step of the sixth step, the PN code timing error is performed by performing the threshold check Ct times on the maximum value of the vector elements of the fast Fourier transform result on the despread signal and passing Tt times or more. It is characterized by determining that does not occur.

Even more preferably, in the second sub-step of the sixth step, a PN code timing error occurs when the threshold check result passes Tt times continuously with respect to the maximum value of the fast Fourier transform result vector elements for the despread signal. It is characterized by determining not to.

Even more preferably, the second sub-step of the sixth step may be performed by performing the approximate Doppler frequency shift estimate Cf times from the position of the maximum value of the fast Fourier transform result vector elements for the despread signal, thereby performing the same Tf times or more. It is characterized by determining that the Doppler frequency shift estimation error is equal to or less than the frequency detection resolution of the fast Fourier transform.

Even more preferably, the second sub-step of the sixth step is performed so that the Doppler frequency is obtained if the approximate Doppler frequency transition estimate is equal to Tf times from the position of the maximum value of the fast Fourier transform result vector elements for the despread signal. It is characterized by determining that the transition estimation error is equal to or less than the frequency detection resolution of the fast Fourier transform.

Even more preferably, in the second sub-step of the sixth step, it is determined that a residual carrier frequency recovery error has occurred in the first sub-step, it is determined that no PN code timing error occurs, and the Doppler frequency shift estimation If it is determined that the error is equal to or less than the frequency detection resolution of the fast Fourier transform, it is determined that the residual carrier frequency error is equal to or less than the frequency detection resolution of the fast Fourier transform and falls within a range that can be restored in the residual carrier frequency precision recovery step. .

Hereinafter, the "Doppler frequency shift compensation method using a fast Fourier transform" according to an embodiment of the present invention with reference to the accompanying drawings in detail as follows.

1 is a flowchart illustrating Doppler transition and pseudo noise (PN) code timing recovery after receiving a radio frequency signal at a receiver of a wireless communication system using a direct sequence spread spectrum signal (DS-CDMA) scheme. 2 is a block diagram illustrating a receiver of a wireless communication system corresponding to the flowchart of FIG. 1.

2, a receiver of this wireless communication system includes a radio frequency amplifier and a filter 201, a radio frequency / intermediate frequency converter 202, an intermediate frequency amplifier and a filter 203, and an intermediate frequency to baseband frequency converter ( 204, an analog-to-digital converter 205, a synchronization unit 210, a demodulator 206, and a synchronization loss detection unit 207. The synchronization unit 210 includes a pseudo noise code tracking unit 211, a despread correlator 212, a mixer, a numerically controlled oscillator 216, a local pseudo noise code generator 215, and a pseudo noise code synchronization acquisition unit 2146. ), And the Doppler frequency shift estimating unit 213.

The signal received by the antenna of the receiver is a mixed signal of a direct sequence spread band signal and a pilot channel signal of several users, and the radio frequency amplifier and the filter 201 amplify the received radio frequency (RF) signal and band The reception channel is selected by removing the external noise (S101). The radio frequency / intermediate frequency converter 202 frequency-converts this radio frequency signal to an intermediate frequency (S102). The intermediate frequency amplifier and the filter 203 amplify the intermediate frequency signal, remove the noise, and filter (S103). The intermediate frequency-baseband frequency converter 204 converts the intermediate frequency signal to the baseband (S104).

After the frequency shift to this baseband, the analog-to-digital converter 205 is sampled at the chip frequency and converted into a digital signal (S105). At this time, the frequency shift to the baseband is performed by a local oscillator and a mixer of a fixed frequency without considering the Doppler frequency shift, and thus there is a residual carrier frequency error due to the Doppler frequency shift in the baseband signal. This error is compensated for by the synchronization unit 210.

The baseband signal sampled at chip frequency is mixed in the mixer with the locally generated pseudonoise code assigned to the pilot channel and despread in despread correlator 212. The pseudo noise (PN) code synchronization acquisition unit 214 acquires the synchronization of the despread pseudo noise (S106), and the Doppler frequency shift estimation unit 213 estimates the approximate Doppler transition of the despread pseudo noise. Compensate (S107). Thereafter, a precise Doppler transition estimation and compensation step S108 and a precision compensation and tracking step S109 of pseudo noise code synchronization are performed simultaneously.

When this series of synchronization operations is completed and the error of the carrier frequency and the pseudo noise code timing converges within an acceptable range, the output signal of the despread correlator 212 in the synchronization unit 210 is transmitted to the demodulator 206 to demodulate the symbols. Perform (S110). In this case, the synchronization of the carrier frequency and the pseudo-noise code timing may be lost while the receiver is in operation due to noise and interference signals included in the received signal. The synchronization loss detection unit 207 observes whether or not signal quality deterioration or synchronization loss occurs in the demodulation process, and the signal quality is degraded (S111). If the pseudo-synchronization code synchronization is lost (S112), the synchronization unit 210 is detected. In step S113, an initialization signal for pseudo noise code synchronization acquisition is output. If synchronization is not lost, but the carrier frequency error exceeds the precise Doppler transition compensation range (S114), the approximate Doppler transition compensation loop is initialized (S116), and if the carrier frequency error does not exceed the precise Doppler transition compensation range, the precise Doppler The transition compensation loop is initialized (S115).

3 is a detailed block diagram illustrating a detailed Doppler transition compensation unit. The input signal sampled at the chip frequency is mixed in the mixer 312 with a single tone signal generated by a Numerically Controlled Oscillator (NCO) 311. The frequency of the output signal of the numerically controlled oscillator 311 is given as 0 at the beginning of operation. When the Doppler frequency shift compensator shown in FIG. 3 operates properly, the frequency of the output signal of the numerically controlled oscillator remains in the input signal. Adjusted to approximate carrier frequency error. Thus, no frequency shift occurs by the NCO 311 and the mixer 312 at the beginning of operation.

The signal passing through mixer 312 is correlated by PN code signal assigned to pilot channel generated by local pseudonoise code generator 302 by correlator 301 of length X chip (X: positive integer). Despread by taking. The timing of the locally generated PN code can be obtained using one of the techniques described in the following documents.

First, U. Cheng, W. J. Hurd and J. I. Statman, Spread-spectrum code acquisition in the presence of Doppler shift and data modulation, IEEE Transactions on Communications, vol. 38, no. 2, pp. 241-250, February 1990.

Second, M. Ishizu, M. Katayama, T. Yamazato and A. Ogawa, Initial acquisition of code timing and carrier frequencies of CDM down-link signals in multiple-LEO-satellite communication systems, IEICE Transactions on Fundamentals, vol. E81-A, no. 11, pp. 2281-2290, November 1998.

Third, P. M. Grant, S. M. Spangenberg, I. Scott, S. McLaughlin, G. J. R. Prvey and D. G. M. Cruickshank, Doppler estimation for fast acquisition in spread spectrum communication systems, Proceedings of IEEE ISSSTA, vol. 1, pp. 106-109, 1998.

The longer the correlator length used for despreading, the lower the noise component power of the correlator output signal, but also the signal component power of the correlator output signal when there is a non-zero residual carrier frequency error.

4 is a graph showing signal component power of a correlator output signal according to the correlator length for various residual carrier frequency errors Δf. From this result, it is preferable that the length of the correlator is selected as the maximum value within a range (for example, 0.1 dB) where the signal power loss in the correlator is sufficiently small. The despreading pilot channel signal becomes a discrete time sinusoidal signal having a frequency corresponding to the residual carrier frequency error.

The despread correlator output signal c [n] (n: integer) is stored in a buffer 303 of length P (P: positive integer) and then sized after P zero-paddings have been made. The fast Fourier transformer 304, which is 2P, is converted into the frequency domain as shown in Equation 1.

Where k = 0, 1, 2P-1.

Since the sampling frequency of the input signal of the fast Fourier transformer 304 is 1 / X times the chip frequency, the frequency resolution of the fast Fourier transform result is 1 / (2XP) times the chip frequency, and the k-th frequency bin and k are zero-padded. Since orthogonality is not established between + 2m-1 (m: integer) th frequency bins, it has a nonzero correlation value. The fast Fourier transformed signal is temporarily stored in the buffer 305 and square-scaled by the magnitude-squared operator 306.

FIG. 5 is a graph illustrating interference states between elements of a fast Fourier transform result vector with respect to a frequency of a single tone input. In FIG. 5, the fast Fourier transform size is set to 4 in consideration of the complexity of the fast Fourier transform processor. Therefore, two correlation results are used as the fast Fourier transform input signal. If the fast Fourier transform size is set to 4 and two zero paddings are used, the fast Fourier transform operator can be simply implemented with four adders and four subtractors, as shown in FIG. .

From the results of FIG. 5, it can be seen that a correlation value is 0 due to orthogonality between two frequency bins separated by even frequency bins and that the correlation value is not 0 between two frequency bins separated by odd frequency bins. Also, in the latter case, it can be seen that the correlation value between adjacent frequency bins is significantly larger than the correlation value between non-adjacent bins. The present invention uses a correlation value between two adjacent frequency bins and uses the symmetry of the correlation characteristics shown in FIG. 5. In this case, since only C [1] and C [3] are needed in FIG. 6A, the fast Fourier transform processor has a simple structure having two adders and two subtractors as shown in FIG. Can be replaced with a processor.

The present invention assumes a case where the residual carrier frequency error is within the range of Equation 2.

In this case, by comparing the magnitudes of C [1] and C [2P-1] in the fast Fourier transform result vector, the residual carrier frequency error of the input signal may be estimated. That is, the present invention estimates the residual carrier frequency error from the difference between the magnitude squared values of C [1] and C [2P-1] in the fast Fourier transform result vector according to equation (3).

That is, the magnitude squared value of C [1] and the magnitude squared value of C [2P-1] output from the magnitude squared operator 306 are stored in the registers 307 and 308, respectively. The difference value is calculated, and the loop filter 310 estimates the residual carrier frequency error from the difference value and provides it to the numerically controlled oscillator 311.

The magnitude squared operation in this equation is described in E. D. Kaplan, Understanding GPS: Principles and Applications. By using the Robertson approximation presented in Boston: Altech House, 1996, it can be replaced with a size operation that can be simply implemented, where Equation 3 can be replaced by Equation 4.

In the present patent, a residual carrier frequency error estimation equation according to Equation 3 will be described. FIG. 7 illustrates the characteristics of the estimation result of the residual carrier frequency error using a fast Fourier transform processor having a structure as shown in FIG. 6 (b) and Equation 3 for a channel environment without background noise according to various correlation lengths. It is a graph figure. From this result, it can be seen that the residual carrier frequency error compensation can be performed for the residual carrier frequency error within the range corresponding to Equation 2.

8 is a block diagram of a PN code timing and a 2D retrieval technique for Doppler frequency transition according to an embodiment of the present invention.

This is suggested by the following documents.

First, U. Cheng, W. J. Hurd and J. I. Statman, Spread-spectrum code acquisition in the presence of Doppler shift and data modulation, IEEE Transactions on Communications, vol. 38, no. 2, pp. 241-250, February 1990.

Second, S. Okuda, M. Katayama, T. Yamazato and A. Ogawa, A new block demodulator for DS / SS signal with carrier frequency offset, Proceedings of IEEE International Symposium on PIMRC, vol. 1, pp. 203-207, 1995.

Third, J. Y. Kim and J. H. Lee, Acquisition performance of a DS / CDMA system in a mobile satellite environment, IEICE Transactions on Communications, vol. E80-B, no. 1, pp. 40-48, January 1997.

Fourth, M. Ishizu, M. Katayama, T. Yamazato and A. Ogawa, Initial acquisition of code timing and carrier frequencies of CDM down-link signals in multiple-LEO-satellite communication systems, IEICE Transactions on Fundamentals, vol. E81-A, no. 11, pp. 2281-2290, November 1998.

Fifth, P. M. Grant, S. M. Spangenberg, I. Scott, S. McLaughlin, G. J. R. Prvey and D. G. M. Cruickshank, Doppler estimation for fast acquisition inspread spectrum communication systems, Proceedings of IEEE ISSSTA, vol. 1, pp. 106-109, 1998.

The two-dimensional retrieval technique of PN code timing and Doppler frequency shift acquisition using fast Fourier transform first changes the timing of the locally generated PN code for despreading the received signal at regular intervals (one chip or half chip interval). As a result of the fast Fourier transform processing on the despread signal, the PN code timing acquisition is performed by finding the timing of the local PN code whose maximum value of power of the elements of the vector exceeds a predetermined threshold.

Also, a rough estimate of the Doppler frequency transition is obtained from the position of the element having the maximum power in the fast Fourier transform processing result vector for the PN code timing thus obtained. In this process, noncoherent combining of length M (M: positive integer) is performed on the complex-squared calculation result to increase the reliability of the estimation result. More than a certain number of times to verify. Here, the Doppler frequency transition estimate is approximated because it cannot have accuracy within the limits of the frequency detection resolution of the fast Fourier transform. In most systems having a realistic complexity, a precision of more than the frequency detection resolution of the fast Fourier transform is required, and thus a residual carrier frequency recovery technique such as the present invention is required. The obtained PN code timing is more precisely restored by a PN code tracking unit performed after Doppler frequency shift compensation or simultaneously with Doppler frequency shift compensation.

Approximate estimates of the Doppler frequency shift are used to adjust the frequency of the locally generated carrier mixed with the input signal. After the approximate Doppler frequency shift compensation is performed, the frequency bin elements C [1] and C [2P-1] adjacent to element C [0] corresponding to frequency 0 in the fast Fourier transform result vector are proposed in the present invention. After the approximate Doppler frequency compensation by the technique, the residual carrier frequency error is detected. The detected residual carrier frequency error is used to reduce the noise component and the high frequency component through the loop filter and to adjust the locally generated carrier frequency with the coarse Doppler frequency shift detection result detected earlier.

Referring to FIG. 8, a Doppler frequency shift compensator using a fast Fourier transform according to the present invention includes a despread correlator 801, a fast Fourier transform processor 804, and a complex magnitude squared operator 809. Share with Therefore, when the present invention is applied to a receiver using a two-dimensional retrieval technique, only the subtractor, the loop filter 818, and the numerically controlled oscillator 820 can perform Doppler frequency shift compensation with minimal complexity.

FIG. 9 is a flowchart illustrating an operation of a 2D search related unit using a fast Fourier transformer of FIG. 8. In the figure, as a verification technique for the PN code timing and the approximate Doppler frequency shift detection result, Tt (Tt: positive integer) or Ct (Ct: positive integer) or Cf (Cf: positive integer) If a Tf (Tf: positive integer) result in the same estimate more than once, the method determines that the estimate is reliable.

Where Ct is the number of threshold checks to be performed for PN code acquisition acquisition verification, Cf is the number of checks to be performed for approximate Doppler transition estimation verification, and Tt is the minimum threshold to pass for PN code synchronization acquisition verification. Tf is the minimum number of inspections that must pass for verification of the approximate Doppler transition estimate.

Referring to FIG. 9, first, the test count Ntest, the test pass count Tcnt in the PN code synchronization acquisition verification, and the test pass count Fcnt in the approximate Doppler transition estimation verification are initialized to 0 (S901). .

The PN code signal assigned to the pilot channel generated by the local pseudo-noise code generator 802 is then despread by correlation being taken by the correlator 801 of length X chip (X: positive integer). The despread correlator output signal c [n] (n: integer) is stored in a buffer 803 of length P (P: positive integer) and then sized after P zero-paddings. The fast Fourier transform 804, which is 2P, is converted into the frequency domain. The fast Fourier transformed signal is temporarily stored in the buffer 805, square-scaled by the magnitude squared operator 806, and is non-phase coupled to the output signal of the shift register 810 by the combiner 811 (S902).

This output signal is input to the detector 812 to detect the maximum power value and its position among the elements of the fast Fourier transform output vector (S903). At this time, if the maximum power value is greater than the threshold (S904), the process proceeds to step S905, and if the maximum power value is not greater than the threshold (S904), the process proceeds to step S915.

In step S905, Tcnt is increased by 1 and the number of times of test (Ntest) is also increased by 1. When the number of times of test (Ntest) is 0 (S907), if it is 0, the position of the maximum power value is determined by the variable INDEXprev of the register 815. After the data has been stored (S908), the process returns to step S902.

On the other hand, if the number of checks is not zero in step S907, the flow advances to step S909 to check whether the position of the maximum power value among the elements of the fast Fourier transform output vector is equal to the variable INDEXprev. If the result of the check in S909 is YES, Fcnt is increased by 1 and the flow proceeds to step S911. If the result of the check in S909 is no, the process proceeds directly to step S911.

If the number of inspections in step S911 is smaller than a predetermined predetermined value A, Tcnt and Tt are compared (S912). If Tcnt is not greater than Tt, the process proceeds to step S916. If Tcnt is greater than Tt, the process proceeds to step S913. In step S913, Fcnt and Ft are compared (S913). As a result of the comparison of step S913, if Fcnt is larger than Ft, the approximate Doppler transition estimation is completed (S914). This approximate Doppler frequency shift estimate is provided to the adder 819 and added to the precise residual carrier frequency error estimate provided from the loop filter 818. If Fcnt is not greater than Ft as a result of the comparison in step S913, the process returns to step S902 and the above steps are repeated.

Meanwhile, in step S916, the clock generator 814 is driven to move the local PN code timing by half a chip period, and then the process proceeds to step S901 to repeat the above steps.

Instead of the verification method shown in FIG. 9, a method of determining that the estimation result is reliable when the estimation results of consecutive Tt or Tf times coincide may be adopted.

10 is a flowchart illustrating a PN code timing and Doppler frequency shift compensation technique using a two-dimensional search and an interworking scheme of the present invention. In the drawings, a method of performing a residual carrier frequency recovery process using the present invention in two steps is illustrated. That is, when the PN code timing and the approximate Doppler frequency transition estimation of FIG. 9 are completed (S1001), the loop bandwidth is relatively widened by shortening the residual carrier frequency error by setting the correlation operation length short in the step before the fast Fourier transform operation. After performing the first precision recovery step (S1002) to reduce within a certain range in time, and then reduce the loop bandwidth by increasing the correlation operation length to reduce the residual carrier frequency error in the steady state within the final target range The second precision restoration step S1003 is performed.

In this case, instead of detecting whether the residual carrier frequency error converged within a predetermined range by the first precision reconstruction step, the bandwidth of the loop filter may be reduced without a separate convergence detection process after the maximum convergence time estimated by the preliminary analysis result. It may be. In the second precision recovery step for the residual carrier frequency error, if the residual carrier frequency error is out of the restorable range of the residual carrier frequency recovery loop by continuously checking whether the residual carrier frequency error is out of the restorable range, it will be described with reference to FIG. 9. The PN code timing and the approximate Doppler frequency shift estimation and verification process are performed, and the procedure is repeated from the appropriate step of FIG.

In this case, as a detection method for the case where the residual carrier frequency error is out of the restorable range, elements C [1], C [2P-1], and C [0 of the fast Fourier transform result vector used for the residual carrier frequency error detection are used. In addition to the detection by the pattern of the complex magnitude values of], it can be performed by measuring the SNR or the bit detection error rate in the demodulator after all synchronization processing including the present invention.

If synchronization is lost during the execution of the second precision restoration step (S1004), the process proceeds to step S1003 again to perform the second precision restoration step again.

Next, to verify the operation of the PN code synchronization acquisition and the approximate Doppler transition compensator (S1005), if the Pt code timing error does not pass more than Tt times out of the Ct threshold check results, it is determined that a PN code synchronization error occurs. It is determined that it is lost (S1006), and the process proceeds to the PN code synchronization reacquisition step (S1007). This step S1007 is associated with step S917 of FIG. 9, and is repeated from S901 of FIG. 9 after changing the locally generated PN code timing at a predetermined interval.

On the other hand, if the PN code timing and the approximate Doppler frequency shift estimation and verification result and the Cf times Doppler frequency shift estimation result do not match at least Tf times, the PN code timing error does not occur and the frequency detection resolution of the fast Fourier transform is higher than that. It is determined that the Doppler frequency error has occurred, that is, it is determined that the Doppler transition acquisition is lost (S1008), and the flow proceeds to the Doppler transition acquisition loss step (S1009). This step S1009 is associated with step S918 in Fig. 9, and is repeated from step S901 in Fig. 9 while maintaining the current PN code timing. If no PN code timing error occurs and the Doppler frequency shift estimation error is verified to not exceed the frequency detection resolution of the fast Fourier transform, the residual carrier frequency error is smaller than the frequency detection resolution of the fast Fourier transform and the residual carrier is After determining that the frequency falls within a range larger than the range that can be restored in the second precision recovery step of the frequency, the fine Doppler transition compensation loop is initialized (S1010), and then repeated from the first precision recovery step of the residual carrier frequency.

In this document, a wireless communication system using a direct sequence spread spectrum signal is provided as an embodiment for effectively explaining the present invention. However, the present invention can be applied to a single frequency signal except for a despreading process for a received signal. .

While the invention has been described above, these examples are intended to illustrate rather than limit this invention. It will be apparent to those skilled in the art that various changes, modifications, or adjustments to the above embodiments are possible without departing from the technical details of the present invention. Therefore, the protection scope of the present invention will be limited only by the appended claims, and should be construed as including all such changes, modifications or adjustment examples.

As described above, according to the present invention, the Doppler frequency shift compensation with low complexity can be compensated by adjusting the output frequency of the local oscillator by comparing the powers of two frequency signal components adjacent to the center frequency in the frequency domain by using a small fast Fourier transform. It has an effect. In particular, in the case of a receiver using a two-dimensional retrieval technique using a fast Fourier transform, there is an advantage that there is almost no increase in complexity additionally required to implement the present invention.

In addition, the present invention can adjust the frequency recovery precision by changing the correlation length in the previous step of the fast Fourier transform operation without changing the fast Fourier transform operator, thereby increasing the initial frequency acquisition speed and at the same time improving the final frequency recovery accuracy. It has an effect.

Claims (16)

A first step of frequency shifting the baseband with respect to the received signal, detecting an approximate Doppler frequency shift, and compensating for the residual carrier frequency of the baseband received signal according to the detection result; Calculating a power distribution of discrete frequency components by performing a fast Fourier transform operation on the approximate Doppler frequency shift detection and compensation signal; Estimating a residual carrier frequency error from the fast Fourier transform calculation result; A fourth step of reducing noise using a low pass filter for the residual carrier frequency error estimation result; A fifth step of compensating for a residual carrier frequency error by adjusting a frequency of a local oscillator according to the low pass filtered result and proceeding to the second step; And The Doppler frequency shift compensation method using the fast Fourier transform, characterized in that for performing the residual carrier frequency error compensation, the sixth step of repeating from the first step or the third step in accordance with the type of error if an error occurs. The method of claim 1, wherein the first step, If the received signal is a direct sequence spread band signal, the frequency shifted to the baseband and then correlated with a locally generated PN code to inversely convert the received signal and detect an approximate Doppler frequency shift. Doppler frequency shift compensation method using. The method of claim 1, wherein the second step, A first substep of deliberately generating correlations between adjacent frequency bins by padding the same number of zeros with the output signal of the first step, and a second substep of performing a fast Fourier transform operation with a magnitude of the zero-padded signal Including, The third step, Using the fast Fourier transform, the residual carrier frequency error is estimated from the signal component difference between two adjacent frequency bins and a frequency bin corresponding to frequency 0 using the symmetry of the correlation between adjacent frequency bins of the fast Fourier transform output vectors. Doppler frequency shift compensation method. The method of claim 3, wherein the second substep is A Doppler frequency shift compensation method using a fast Fourier transform, characterized by using a fast Fourier transform calculator of size 4 consisting of two adders and two subtractors. The method of claim 3, wherein the third step, Using the fast Fourier transform, the residual carrier frequency error is estimated by applying the C [1] and C [2P-1] of the fast Fourier transform operation vector to the following equation to obtain the difference between the magnitude squared values. Doppler frequency shift compensation method. [Equation] The method of claim 3, wherein the third step, Doppler frequency shift using fast Fourier transform by estimating the residual carrier frequency error by applying the magnitude difference by applying C [1] and C [2P-1] to the following equation Compensation Method. [Equation] The method of claim 1, wherein the fourth step, A first substep of passing the approximate Doppler frequency shift compensated signal through a wide bandwidth loop filter to quickly converge residual carrier frequency error within a predetermined range within a short time; And a second sub-step for narrowing the loop filter bandwidth to converge the residual carrier frequency error in a steady state within a target tolerance range when the residual carrier frequency error converges within a predetermined range. Doppler frequency shift compensation method. 8. The method of claim 7, wherein the fourth step comprises performing the second sub-step when the residual carrier frequency error is detected and converged within a predetermined range, and performing the second sub-step. . 8. The method of claim 7, wherein the fourth step comprises performing the second sub-step after forcibly resetting the bandwidth of the loop filter after a time estimated that the residual carrier frequency error has converged within a predetermined range. Doppler frequency shift compensation method using fast Fourier transform. The method of claim 2, wherein the sixth step is A first substep of generating a residual carrier frequency recovery error when the residual carrier frequency error is compensated for; A second substep of checking whether an error occurs in PN code timing and Doppler frequency shift compensation for the occurrence error, and A third sub-step of repeating the subsequent steps by proceeding to at least one of a PN code timing acquisition step, an approximate Doppler frequency shift detection step, and a residual carrier frequency error estimation step according to the identified occurrence type of error; Doppler frequency shift compensation method using a fast Fourier transform, characterized in that. The method of claim 8, wherein the first substep is The fast Fourier transform of the fast Fourier transform result vector used for detecting the residual carrier frequency error is detected using a pattern of complex magnitude values of C [1], C [2P-1], and C [0]. Doppler frequency shift compensation method using. The method of claim 8, wherein the second substep is Doppler using Fast Fourier Transform characterized in that the PN code timing error does not occur if Tt is passed more than Tt by performing threshold checking on the maximum value of the vector elements of the fast Fourier transform result of the despread signal. Frequency shift compensation method. The method of claim 8, wherein the second substep is Doppler frequency shift using Fast Fourier Transform characterized in that it is determined that PN code timing error does not occur when the threshold test result passes Tt times with respect to the maximum value of fast Fourier transform result vector elements for despread signal. Compensation Method. The method of claim 8, wherein the second substep is If the fast Fourier transform result of the fast Fourier transform on the despread signal is approximately Cf times the Doppler frequency shift estimation from the position of the maximum value, the Doppler frequency shift estimation error is less than the frequency detection resolution of the fast Fourier transform. Doppler frequency shift compensation method using a fast Fourier transform characterized in that the determination. The method of claim 8, wherein the second substep is If the roughly Doppler frequency shift estimation results in the same Tf times from the position of the maximum value of the fast Fourier transform result vector elements for the despread signal, the Doppler frequency shift estimation error is determined to be equal to or less than the frequency detection resolution of the fast Fourier transform. Doppler frequency shift compensation method using fast Fourier transform. The method of claim 8, wherein the second substep is If it is determined that a residual carrier frequency recovery error has occurred in the first sub-step, it is determined that no PN code timing error has occurred, and it is determined that the Doppler frequency shift estimation error is less than or equal to the frequency detection resolution of the fast Fourier transform, the residual carrier frequency A method for compensating the Doppler frequency shift using a fast Fourier transform, wherein the error is determined to be equal to or less than a frequency detection resolution of the fast Fourier transform and is within a range that is more than a range recoverable in the residual carrier frequency precision restoring step.