JP2012173153A - Frequency tracking device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To calculate a frequency drift quantity with high precision by calculating the frequency drift quantity from a difference of square root of integrated power calculated wit a different coherent addition time.SOLUTION: A frequency tracking device includes an integrator 17 which calculates integrated power |y|through coherent addition of a desired satellite signal (y) with an addition time Mand N-times non-coherent addition; an integrator 18 which calculates integrated power |y|through coherent addition of the satellite signal (y) with an addition time Mdifferent from the addition time Mand N-times non-coherent addition; and a frequency drift computing element 20 which calculates a frequency drift quantity Δffrom the difference between the square root |y| of the integrated power calculated by the integrator 17 and the square root |y| of the integrated power calculated by the integrator 18, and tracks the desired satellite signal (y) based upon the frequency drift quantity Δfcalculated by the frequency drift computing element 20.

Description

  The present invention relates to a frequency tracking device that tracks the carrier frequency of a satellite signal.

  For example, satellite signal receivers such as GPS (Global Positioning System) and Galileo capture satellite signals and determine the position and speed of ground receivers such as vehicles and pedestrians. Since the positioning result (recipient's position and speed) thus obtained is used in a car navigation system, ITS (Intelligent Transportation Systems), etc., high reliability is required. However, when the receiver is located in an environment where the signal strength of the satellite is easily deteriorated as surrounded by trees, it is difficult to track the carrier frequency of the satellite signal with high reliability. If the tracking of the carrier frequency becomes difficult, the accuracy of the positioning result is deteriorated or the positioning is not performed. Even in such a case, the satellite signal receiving device includes a frequency tracking device in order to highly track the carrier frequency of the satellite signal.

  The frequency tracking device calculates the amount of frequency deviation between the local frequency signal of the desired satellite (the local signal converted into a predetermined frequency that can be handled easily by the baseband processing unit) and the replica signal generated in the device. The carrier NCO (Numerically Controlled Oscillator) corrects the frequency of the replica signal so as to minimize the frequency deviation based on the frequency deviation, and repeats the correction to synchronize the local signal of the desired satellite and the replica signal. It tracks the carrier frequency of satellite signals. As shown in FIG. 3, the conventional frequency tracking device 100 is known to include a phase rotator 101, an integrator 102, a discriminator 103, a loop filter 104, an edge information generator 105, and a carrier NCO 106. The frequency tracking device 100 performs complex multiplication of the local signal of the desired satellite and the replica signal by the phase rotator 101, and performs integration (filtering) by the accumulator 102 in order to suppress high frequency noise superimposed on the complex multiplied signal. The frequency shift amount between the local signal of the desired satellite and the replica signal is calculated by the nominator 103, and noise superimposed on the frequency shift amount is further suppressed by the loop filter 104. Based on the frequency shift amount in which the noise is suppressed, the carrier NCO 106 Correct the frequency of the replica signal. In this way, control is performed so that the frequency deviation between the frequency of the local signal of the desired satellite and the frequency of the replica signal becomes zero. The edge information generator 105 notifies the accumulator 102 of the navigation data superimposed on the local signal of the desired satellite and the sign inversion information (edge information) based on the secondary code. For example, the sign reversal of GPS L1C / A navigation data appears where there is a possibility of sign reversal every 20 ms because the baud rate of the navigation data is 50 bps. The accumulator 102 performs integration so as not to cross the sign inversion boundary, thereby obtaining the maximum gain integration result.

  Incidentally, it is known that the satellite carrier phase (hereinafter referred to as “phase”) becomes unstable in an environment where the signal strength of the satellite is likely to deteriorate. For example, an obstacle such as a leaf suddenly interrupts between the satellite and the receiver, whereby the satellite signal is refracted and reflected, thereby causing a sudden offset change and cycle slip in the phase, and the phase becomes unstable. Since the discriminator 103 calculates the frequency shift amount by differentiating the phase in discrete time, if the phase is unstable, the calculated frequency shift amount includes an error due to phase instability, For this reason, there is a possibility that frequency tracking becomes unstable or an incorrect frequency is tracked.

  The technique of Patent Document 1 is known as a frequency estimation device (frequency tracking device) that prevents an erroneous frequency from being tracked even if signal intensity or noise varies with time. This frequency estimation apparatus uses two or more reception channels (reception circuits) to obtain an estimation frequency (frequency deviation amount), obtains an evaluation frequency, and obtains a frequency deviation amount from the obtained two or more evaluation frequencies. Is calculated. The reception channel is composed of a circuit equivalent to the conventional frequency tracking device 100 shown in FIG. 3, and when tracking the frequency of one satellite, two or more reception channels are used. There was a drawback that the circuit scale and current consumption more than doubled were required. Moreover, the technique of patent document 2 is known as a technique regarding the satellite signal tracking apparatus (frequency tracking apparatus) which suppressed the increase in circuit scale and consumption current. Since this satellite signal tracking device obtains the amount of frequency deviation by the signal power (integrated power) ratio, the frequency deviation (frequency shift) is such that the integrated power of the numerator and the integrated power of the denominator are equivalent. When this occurs, there is a drawback that the amount of frequency deviation cannot be accurately detected by calculating the ratio of integrated power.

Patent No. 3924500 JP 2008-111684 A

  In view of this, the present invention calculates the frequency shift amount based on the difference between the square roots of the integrated powers calculated with different coherent addition times in order to minimize the increase in circuit scale and power consumption and accurately detect the frequency shift amount. An object of the present invention is to provide a frequency tracking device capable of calculating a highly accurate frequency shift amount.

  In order to solve the above-mentioned problem, the invention according to claim 1 is characterized in that a desired satellite signal and a replica signal from a carrier NCO are multiplied by a complex number by a phase rotator, and the signal is accumulated by an accumulator based on the complex number multiplication value. A frequency tracking device that calculates a frequency shift amount and feeds back the frequency shift amount so that a frequency shift amount between the carrier frequency of the satellite signal and the frequency of the replica signal becomes zero. A first integrator for coherently adding the desired satellite signal for a first predetermined addition time and performing a first predetermined number of noncoherent additions to calculate a first integrated power; and the desired satellite The signals are coherently added at a second predetermined addition time different from the first predetermined addition time, and a second predetermined number of non-coherent additions are performed. A second integrator that calculates the integrated power of the first integrated power, and a frequency shift calculator that calculates a frequency shift amount from the difference between the square root of the first integrated power and the square root of the second integrated power, A desired satellite signal is tracked based on the frequency shift amount calculated by the frequency shift calculator.

  According to the present invention, the amount of frequency deviation is calculated from the difference between the square root of the first integrated power calculated by the first integrator and the square root of the second integrated power calculated by the second integrator. Is done.

  According to a second aspect of the present invention, in the frequency tracking device according to the first aspect, a discriminator that calculates a frequency shift amount by performing a discrete time differentiation of a phase shift amount of the desired satellite signal, and the frequency shift And a weighted averager that calculates an optimal frequency shift amount based on the frequency shift amount calculated by the arithmetic unit and the frequency shift amount calculated by the discriminator.

  According to the first aspect of the present invention, the frequency deviation amount is calculated by calculating the square root of the first integrated power obtained by coherent addition in the first predetermined addition time and the second predetermined addition time different from the first predetermined addition time. Since the calculation is performed from the difference from the square root of the second integrated power obtained by coherent addition, an offset is not added to the calculated frequency shift amount. That is, since the first predetermined addition time and the second predetermined addition time are different, the frequency deviation amount can always be calculated from the difference between the square root of the first integrated power and the square root of the second integrated power. Since the calculated frequency deviation amount does not use a calculation for differentiating the phase deviation amount in discrete time as in the case of a conventional discriminator, even if sudden phase change and cycle slip occur frequently. The amount of frequency deviation can be obtained with high accuracy.

  According to the second aspect of the present invention, the weighted averager can determine the reception state of the satellite signal based on the frequency shift amount calculated by the frequency shift calculator and the frequency shift amount calculated by the discriminator. Accordingly, an appropriate frequency shift amount can be calculated.

It is a schematic block diagram which shows the frequency tracking apparatus which concerns on embodiment of this invention. It is a figure which shows the relationship between the frequency shift amount computed with the 1st integrator and the 2nd integrator in the frequency tracking apparatus of FIG. 1, and the deterioration degree of integrated power. The relationship between the frequency shift amount and the deterioration degree of the integrated power is a sinc function. It is a schematic block diagram which shows the conventional frequency tracking apparatus.

  The present invention will be described below based on the illustrated embodiments.

  1 and 2 show an embodiment of the present invention.

A plurality of frequency tracking devices 1 are provided in satellite signal receiving devices (not shown) such as GPS and Galileo. That is, the frequency tracking device 1 is provided for each of a plurality of satellite signals received by the satellite signal receiving device, and each frequency tracking device 1 tracks a carrier frequency F ₁ of a different desired satellite signal y. The satellite signal y is received by the antenna (not shown) of the satellite signal receiving device and transmitted to the frequency tracking device 1. The frequency tracking device 1 generates a replica signal of the satellite signal y, thereby targeting the satellite signal y. As shown in FIG. 1, the frequency tracking device 1 mainly includes a phase rotator 11, a carrier NCO 12, an edge information generator 13, integrators 14, 17, 18, 19, a discriminator (frequency discriminator) 15, a loop. A filter 16, a frequency shift calculator 20, and a weighted averager 31 are provided. This frequency tracking device 1 multiplies a desired satellite signal y (local signal) and a replica signal from a carrier NCO 12 by a complex number by a phase rotator 11 and accumulates a complex number multiplication value by accumulators 14, 17, 18, and 19, The frequency shift amounts Δf ₁ and Δf ₂ are calculated from the integrated values by the discriminator 15 and the frequency shift calculator 20, and the frequency shift amount Δf ₃ is calculated by the weighted averager 31 and fed back to the carrier NCO 12. frequency shift amount of the frequency F ₂ of the carrier frequencies F ₁ and the replica signal of the signal y are used to control so as to zero. Here, the satellite signal y input to the frequency tracking device 1 can be expressed by the following equation using a frequency shift amount Δf and a code error Δτ, if noise is ignored.

  In equation (1), the code error Δτ becomes almost 0 after the satellite signal y is captured. Therefore, if the code correlation function R (Δτ) = 1 and the navigation data D is ignored, the satellite signal y can be expressed by the following equation.

Further, in order to remove the phase difference, the norm of the satellite signal y is calculated.

  The phase rotator 11 outputs I and Q signals (Inphase and Quadrature signals) IQ generated by complex multiplication of the carrier wave of the satellite signal y and the replica signal to the integrators 14, 17, 18 and 19.

The carrier NCO 12 generates a replica signal having a frequency F ₂ that is substantially the same as the carrier frequency F ₁ of the desired satellite signal y based on a frequency deviation amount Δf ₃ described later, and outputs the replica signal to the phase rotator 11.

  The edge information generator 13 generates edge information E for obtaining a maximum gain when coherently adding the satellite signal y, and outputs the generated edge information E to the integrators 14, 17, and 18. Here, the edge information E indicates the sign inversion timing based on the navigation data or the secondary code, and the integrators 14, 17 and 18 use this edge information E to perform coherent addition so as not to include the code inversion timing. As a result, the maximum gain can be obtained. The edge information E becomes known information once the sign inversion timing which is the bit boundary of the data bits is detected, and thereafter it is not necessary to detect the sign inversion timing. The sign inversion timing can also be calculated from the ephemeris and positioning information if the parameters (ephemeris) for obtaining the orbit of the satellite have been acquired and positioning has been performed. In the case of Galileo, since the E1-B / C signal has the same chip cycle and the same data bit synchronization of the secondary code, and is synchronized, if the signal is captured, the sign inversion timing of the secondary code from the captured chip. Can be calculated.

  The integrator 14 as the third integrator calculates a discrete time derivative of the phase shift amount between the I and Q signals IQ output from the phase rotator 11 and the desired satellite signal y, and outputs it to the discriminator 15. .

  The discriminator 15 calculates the frequency shift amount and the phase shift amount from the desired satellite signals y and I and the Q signal IQ, and outputs the calculated various shift amounts to the loop filter 16.

The loop filter (low-pass filter) 16 filters noise superimposed on the frequency deviation amount output from the discriminator 15 and outputs the frequency deviation amount Δf ₁ to the weighted averager 31. Further, the correction direction (sign information) C of the frequency shift amount Δf ₁ is calculated from the various shift amounts of the discriminator 15 and output to the calculator 26.

Integrator 17 as the first integrator, coherent addition time of the desired satellite signal y (first predetermined addition time) coherently summed with M _A, non-coherent summation number of (first predetermined number) N _A Non-coherent addition is performed to calculate the first integrated power. That is, the integrator 17, and coherent addition in a coherent addition time M _A on the basis of the edge information E output from the edge information generator 13, the I, Q correlation value and a 1ms correlation value of the satellite signal y, Non-coherent addition is performed for the number N _{A of} non-coherent additions, and integrated power (first integrated power) | y _A | ² is calculated. The integrator 17 outputs the calculated integrated power | y _A | ² to the square root calculator 21. Here, coherent addition is to add the I and Q signals IQ as they are. The non-coherent addition is to add the absolute values of the I and Q signals IQ or the signal power (I ² + Q ² , (I ² + Q ² ) ^1/2 ) of a signal obtained by coherent addition of these.

Accumulator as the second integrator 18, coherent addition time of the desired satellite signal y M _A different coherent addition time (second predetermined addition time) and coherent addition in M _B, non-coherent addition count (second N _B non-coherent addition is performed to calculate integrated power (second integrated power) | y _B | ² . That is, the integrator 18, and coherent addition in a coherent addition time M _B based and edge information E output from the edge information generator 13, the I, Q correlation value and a 1ms correlation value of the satellite signal y, non coherently summed by non-coherent accumulation count _{N B,} the integrated power _| calculating a ^{2 |} y B. The accumulator 18 outputs the calculated accumulated power | y _B | ² to the square root calculator 22.

Here, the integrated powers | y _A | ² and | y _B | ² output from the integrators 17 and 18 are calculated by the following equations.

  The accumulator 19 performs non-coherent addition of I and Q correlation values, which are 1 ms correlation values of the satellite signal, N times and outputs the result to the square root calculator 24.

The frequency shift calculator 20 includes a difference (degradation degree) between the square roots | y _A | and | y _B | of the integrated power output from the integrators 17 and 18, the information Pc output from the integrator 19, and a loop filter. The frequency shift amount is calculated from the frequency shift code information C output from 16. The frequency shift calculator 20 mainly includes square root calculators 21, 22, 24, an adder 23, a low-pass filter 25, and a calculator 26.

The square root calculator 21 calculates the norm y _{A of the} satellite signal and outputs it to the adder 23. From equation 3 and equation 4, norm y _A is expressed by the following equation.

The square root calculator 22 calculates the norm y _{B of the} satellite signal and outputs it to the adder 23. From equation 3 and equation 5, norm y _B is expressed by the following equation.

Here, the coherent addition time _T A ₌ M _A, a T _B = _M B.

The adder 23 calculates the difference between the norm (square root) of the integrated power | y _A | ² calculated by the square root calculators 21 and 22 and the norm of the integrated power | y _B | ² , and outputs the difference to the calculator 26. That is, the adder 23 calculates the coherent addition times T _A and T _{B in the} equations 6 and 7, so that the sincs output from the integrators 17 and 18 are output as shown in FIG. It is the difference in the waveform of the function (the degree of deterioration of the square root of the integrated power, | y _A | − | y _B |). Here, as shown in FIG. 2, for example, when the degree of deterioration of the square root of the integrated power is about 3.6 dB (= −1.4 − (− 5)), the frequency shift amount is about 28 Hz or about −28 Hz. The code information of the frequency shift amount is unknown. For this reason, the adder 23 outputs the absolute value of the frequency shift amount to the calculator 26.

  The square root calculator 24 calculates the norm of the information Pc output from the accumulator 19 and outputs it to the low pass filter 25. Normalization information S output from the low-pass filter 25 is expressed by the following equation.

In equation (8), since coherent addition is not performed, deterioration of the integrated value due to frequency shift is negligible.

  The low-pass filter 25 filters the noise of the normalized information S output from the square root calculator 24 and outputs the filtered noise to the calculator 26.

The computing unit 26 calculates a frequency deviation amount Δf from the deterioration P ₁ of the square root of the integrated power output from the adder 23, the normalized information S output from the low-pass filter 25, and the code information C output from the loop filter 16. ₂ is calculated. First, the deterioration degree P ₁ (| y _A | − | y _B |) can be expressed by the following equation from Equation 6, Equation 7, and Equation 8.

Thereby, the frequency deviation amount Δf is obtained by the following equation.

Further, by adding the code information C, the frequency shift amount Δf ₂ is obtained.

Here, was ^{α = 1/6 · π 2} (T B -T A) (T B + T A). Since T _A and T _B are known information, α can be calculated in advance. Moreover, the following approximate expression was used for Formula 9. ^{sincθ≡sinθ / θ ≒ 1-θ 2} /6 This equation is an equation obtained by Taylor expansion of sinθ at about 0. The code information C is a value indicating the correction direction of the frequency shift amount Δf, and the value is +1 or −1.

The weighted average unit 31 calculates an optimal frequency shift amount based on the frequency shift amount Δf ₂ calculated by the frequency shift calculator 20 and the frequency shift amount Δf ₁ calculated by the discriminator 15. A weighted average of the frequency shift amounts Δf ₂ and Δf ₁ is calculated. Then, the calculated frequency shift amount Δf ₃ is fed back to the carrier NCO 12. Here, when the cycle slip occurs frequently, the weight of the weighted averager 31 is such that the frequency deviation amount Δf ₁ calculated by the discriminator 15 has deteriorated in accuracy, so that the frequency deviation of the discriminator 15 is reduced. The weight of the amount Δf ₁ is reduced, and the weight of the frequency shift amount Δf ₂ calculated by the frequency shift calculator 20 is increased. If the frequency deviation Δf ₁ is as small as several Hz or less and the phase does not become unstable, it is better to use the discriminator 15 for which the instantaneous frequency deviation is obtained by discrete time differentiation. Therefore, the weight of the frequency shift amount Δf ₁ calculated by the discriminator 15 is increased, and the weight of the frequency shift amount Δf ₂ calculated by the frequency shift calculator 20 is decreased.

  In this way, the frequency tracking device 1 tracks the satellite signal of the tracking target (tracking channel). Then, the satellite signal receiving apparatus calculates the user position based on the tracking information, satellite data, user approximate position, and reception time of each tracking channel tracked in this way.

  Next, how to start frequency tracking of the frequency tracking device 1 will be described.

  First, the power of the satellite signal receiving device and the frequency tracking device 1 is turned on, and the satellite signal y received by the antenna of the satellite signal receiving device is searched. When the satellite signal y is captured by the search, the captured information is distributed for each tracking channel in order to track the captured satellite signal y, and the frequency tracking device 1 corresponding to the tracking channel starts frequency tracking as described above. To do.

As described above, according to the frequency tracking device 1, the frequency shift computing unit 20 uses the difference (degradation degree) of the square roots | y _A | and | y _B | of the integrated power calculated by the different coherent addition times M _A and M _B. ) since calculates the frequency shift amount Delta] f ₂ from the phase can be calculated with high precision frequency shift amount Delta] f ₂ even unstable. That is, since the coherent addition times M _A and M _B are different, the frequency shift amount Δf ₂ can always be calculated from the degree of deterioration of the square roots | y _A | and | y _B | of the integrated power. For this reason, in the case where sudden phase offset changes and cycle slips occur frequently, the frequency deviation amount Δf ₁ output from the discriminator 15 is calculated from the discrete time derivative of the phase deviation amount, so the accuracy is reduced. However, since the frequency shift amount Δf ₂ does not use the calculation of differentiating the phase shift amount in discrete time, the frequency tracking can be continued with high accuracy by the frequency shift amount Δf ₂ .

The frequency shift amount Δf ₃ is weighted and averaged by the weighted averager 31 based on the frequency shift amount Δf ₂ calculated by the frequency shift calculator 20 and the frequency shift amount Δf ₁ calculated by the discriminator 15. Since the calculation is performed, the optimum frequency deviation amount Δf ₃ can be calculated according to the reception state of the satellite signal y. In other words, regarding the frequency shift amount used for control, information of two systems having different characteristics is used, so that the frequency tracking performance is improved.

Although the embodiment of the present invention has been described above, the specific configuration is not limited to the above embodiment, and even if there is a design change or the like without departing from the gist of the present invention, Included in the invention. For example, the Kalman filter may be used instead of the weighted averager 31 in order to calculate the optimum frequency deviation amount Δf ₃ .

1 Frequency tracking device 11 Phase rotator 12 Carrier NCO
13 Edge information generator 14 Accumulator (third accumulator)
15 Discriminator 17 Accumulator (first accumulator)
18 Accumulator (second accumulator)
19 accumulator 20 frequency deviation calculator 26 calculator 31 weighted averager Δf frequency deviation amount F ₁ carrier frequency F ₂ frequency of replica signal | y _A | ² first integrated power | y _B | ² second integrated power S normalization information M _a coherent addition time (first predetermined addition time)
M _B coherent addition time (second predetermined addition time)
N _A non-coherent accumulation count (first predetermined number)
N _B non-coherent accumulation count (second predetermined number)

Claims (2)

Complex multiplication of a desired satellite signal and a replica signal from the carrier NCO by a phase rotator, and after integrating the signal by an accumulator based on the complex number multiplication value, calculating a frequency deviation amount and feeding back the frequency deviation amount According to the frequency tracking device for controlling the carrier frequency of the desired satellite signal and the frequency of the replica signal to be zero so that the amount of frequency deviation is zero,
A first integrator for coherently adding the desired satellite signal at a first predetermined addition time and performing a first predetermined number of non-coherent additions to calculate a first integrated power;
A second integration for calculating a second integrated power by coherently adding the desired satellite signal with a second predetermined addition time different from the first predetermined addition time and performing a second predetermined number of non-coherent additions. And
A frequency shift calculator that calculates a frequency shift amount from the difference between the square root of the first integrated power and the square root of the second integrated power;
Based on the frequency shift amount calculated by the frequency shift calculator, a desired satellite signal is tracked.
A frequency tracking device. A discriminator for calculating a frequency shift amount by differentiating a phase shift amount of the desired satellite signal in a discrete time;
A weighted averager that calculates an optimal frequency shift amount based on the frequency shift amount calculated by the frequency shift calculator and the frequency shift amount calculated by the discriminator;
The frequency tracking device according to claim 1, further comprising: