The invention relates to a phase frequency detection device that is mounted on, for example, a signal measuring device, etc., and that detects a phase and a frequency of an input digital signal.
A phase frequency detection device disclosed in Patent Literature 1 mentioned later is constituted by A/D conversion circuitry, a digital BPF, delay compensation circuitry, Hilbert transformation circuitry, phase angle computation circuitry, a memory, and frequency computation circuitry.
The A/D conversion circuitry converts an input analog signal into a digital signal and outputs the digital signal to the digital BPF.
The digital BPF outputs only a digital signal in a desired frequency band among digital signals output by the A/D conversion circuitry to the delay compensation circuitry and the Hilbert transformation circuitry.
After receiving the digital signal of the desired frequency band from the digital BPF, the delay compensation circuitry performs a time-shift of the digital signal by a time corresponding to a delay time caused by a Hilbert transformation in the Hilbert transformation circuitry.
After receiving the digital signal of the desired frequency band from the digital BPF, the Hilbert transformation circuitry performs a Hilbert transformation on the digital signal, and thereby outputs a signal differing from the digital signal in a phase by 90 degrees.
Note that the higher the accuracy required for the amplitude and phase of an output signal from the Hilbert transformation circuitry, the longer the delay time resulting from a Hilbert transformation. Therefore, the number of stages of delay circuitry included in the delay compensation circuitry increases, resulting in several tens to hundreds of stages.
The phase angle computation circuitry computes a phase of the input analog signal by performing an arctangent computation based on the output signal of the delay compensation circuitry and the output signal of the Hilbert transformation circuitry, and outputs the resulting phase to the memory and the frequency computation circuitry.
The frequency computation circuitry computes a frequency of the input analog signal by using the phase output by the phase angle computation circuitry and a phase obtained at one sampling time before, which has been stored in the memory.
Patent Literature 1: JP 2005-91255 A (e.g., paragraph [0008] and FIG. 1)
Since the conventional phase frequency detection device is constituted in the above-described manner, the higher the accuracy required for the amplitude and phase of an output signal from the Hilbert transformation circuitry, the longer the delay time resulting from a Hilbert transformation, resulting in increase in the number of stages of delay circuitry of the delay compensation circuitry. Hence, in a case where a signal to be input is a short-pulse signal, there may cause a problem that the phase and the frequency of the input signal are hard to be detected. In addition, there may be another problem that, as the number of stages of delay circuitry increases, the circuitry size increases and also the power consumption resulting from digital computation increases.
The invention is made to solve problems described above. An object of the invention is to obtain a phase frequency detection device that is capable of detecting a phase and a frequency even if a signal to be input is a short-pulse signal, and also capable of suppressing increase in circuitry size and increase in power consumption caused by increase in the number of stages of delay circuitry.
A phase frequency detection device according to the invention includes: filter circuitry including a plurality of transversal filters which are connected in parallel to each other, the plurality of transversal filters changing amplitude and a phase of an input digital signal and outputting different digital signals as respective resulting digital signals whose amplitude and phase have been changed; and a phase frequency computer to compute a phase and a frequency of the input digital signal by performing phase computation and frequency computation using the digital signals output by the plurality of transversal filters, wherein the filter circuitry is constituted such that a first transversal filter, a second transversal filter, and a third transversal filter, as the plurality of transversal filters, are connected in parallel to each other, and the phase frequency computer includes: first division computation circuitry to perform a division computation between a digital signal output by the first transversal filter and a digital signal output by the second transversal filter, and to output a first division computation signal being a result of the division computation; second division computation circuitry to perform a division computation between a digital signal output by the third transversal filter and the digital signal output by the second transversal filter, and to output a second division computation signal being a result of the division computation; multiplication computation circuitry to perform a multiplication computation between the first division computation signal output by the first division computation circuitry and the second division computation signal output by the second division computation circuitry, and to output a multiplication computation signal being a result of the multiplication computation; root computation circuitry to perform a root computation on the multiplication computation signal output by the multiplication computation circuitry, and to output a root computation signal being a result of the root computation; phase computation circuitry to compute a phase of the input digital signal by using the root computation signal output by the root computation circuitry and a sign of the first division computation signal or the second division computation signal; and frequency computation circuitry to compute a frequency of the input digital signal by using the phase computed by the phase computation circuitry.
According to the invention, it is constituted such that filter circuitry including a plurality of transversal filters which are connected in parallel to each other, the plurality of transversal filters changing amplitude and a phase of an input digital signal and outputting different digital signals as respective resulting digital signals whose amplitude and phase have been changed; and a phase frequency computer to compute a phase and a frequency of the input digital signal by performing phase computation and frequency computation using the digital signals output by the plurality of transversal filters. Thus, there are advantageous effects that, even if a digital signal to be input is a short-pulse signal, a phase and a frequency thereof can be detected, and increase in circuitry size and increase in power consumption caused by increase in the number of stages of delay circuitry can be suppressed.
In order to describe the invention in more detail, modes for carrying out the invention will be described below with reference to the accompanying drawings.
In
The transversal filter 10 as a first transversal filter is constituted by subtraction-type first-order transversal filters 11 and 12 as first subtraction-type first-order transversal filters D, which are connected in series with each other.
The transversal filter 20 as a second transversal filter is constituted such that a subtraction-type first-order transversal filter 21 as a first subtraction-type first-order transversal filter D is connected in series with an addition-type first-order transversal filter 22 as a first addition-type first-order transversal filter I.
The transversal filter 30 as a third transversal filter is constituted by addition-type first-order transversal filters 31 and 32 as first addition-type first-order transversal filters I, which are connected in series with each other.
A phase frequency computer 4 is constituted by division computation circuitries 41 and 42, multiplication computation circuitry 43, n-th root computation circuitry 44, phase computation circuitry 45, and frequency computation circuitry 46. The phase frequency computer 4 performs phase computation and frequency computation that use the digital signals X1, X2, and X3 output by the transversal filters 10, 20, and 30, and thereby computes a phase θX and a frequency fX of the input digital signal Xin.
The division computation circuitry 41 is a first division computation circuitry that divides the output digital signal X2 of the transversal filter 20 by the output digital signal X1 of the transversal filter 10, and outputs, to the multiplication computation circuitry 43, a division computation signal X2/X1 being a result of the division computation as a first division computation signal.
The division computation circuitry 42 is a second division computation circuitry that divides the output digital signal X2 of the transversal filter 20 by the output digital signal X3 of the transversal filter 30, and outputs, to the multiplication computation circuitry 43, a division computation signal X2/X3 being a result of the division computation as a second division computation signal.
The multiplication computation circuitry 43 performs a multiplication computation between the division computation signal X2/X1 output by the division computation circuitry 41 and the division computation signal X2/X3 output by the division computation circuitry 42, and outputs, to the n-th root computation circuitry 44, a multiplication computation signal Y being a result of the multiplication computation.
The n-th root computation circuitry 44 is a root computation circuitry that performs a root computation being an n-th root computation on the absolute value of the multiplication computation signal Y output by the multiplication computation circuitry 43, and outputs, to the phase computation circuitry 45, an n-th root computation signal N being a result of the n-th root computation.
The phase computation circuitry 45 computes a phase θX of the digital signal Xin by performing an arctangent computation on the n-th root computation signal output by the n-th root computation circuitry 44, and performing sign conversion on the result of the arctangent computation depending on the sign of the division computation signal X2/X3 output by the division computation circuitry 42.
Note that the n-th root computation circuitry 44 and the phase computation circuitry 45 may be constituted by, for example, combining a multiplier and a divider, etc. Alternatively, they may be constituted by, for example, a memory or a table that temporarily stores intermediate results of an n-th root computation and an arctangent computation, and a computation circuitry such as a processor performing the above-described computations.
The frequency computation circuitry 46 computes a frequency fX of the input digital signal Xin by using the phase θX computed by the phase computation circuitry 45.
Specifically,
In addition,
Note that, in the example of
The first subtraction-type first-order transversal filter D of
The splitter circuitry 51 is a first splitter circuitry that splits an input digital signal Xin into two.
The one-sampling time delaying circuitry 52 is a first delaying circuitry that delays one digital signal Xin split by the splitter circuitry 51 by one sampling time Ts (=1/fs).
The subtraction computation circuitry 53 subtracts the digital signal, which has been delayed by the one-sampling time delaying circuitry 52, from the other digital signal Xin split by the splitter circuitry 51.
The second subtraction-type first-order transversal filter −D of
The subtraction computation circuitry 54 subtracts the other digital signal Xin split by the splitter circuitry 51 from a digital signal having been delayed by the one-sampling time delaying circuitry 52.
Note that the difference between the first subtraction-type first-order transversal filter D of
The first addition-type first-order transversal filter I of
The splitter circuitry 61 is a second splitter circuitry that splits an input digital signal Xin into two.
The one-sampling time delaying circuitry 62 is a second delaying circuitry that delays one digital signal Xin split by the splitter circuitry 61 by one sampling time Ts.
The addition computation circuitry 63 adds the digital signal delayed by the one-sampling time delaying circuitry 62 to the other digital signal Xin split by the splitter circuitry 61.
The second addition-type first-order transversal filter −I of
The addition computation circuitry 64 receives a digital signal obtained by reversing a sign of the digital signal having been delayed by the one-sampling time delaying circuitry 62, and also receives a digital signal obtained by reversing a sign of the other digital signal Xin split by the splitter circuitry 61. After that, the addition computation circuitry 64 adds the received former digital signal to the received latter digital signal Xin.
Note that the difference between the first addition-type first-order transversal filter I of
Next, operation will be described.
It is assumed that a digital signal Xin[n·Ts] is input to the transversal filters 10, 20, and 30 of the filter circuitry 1 at a discrete time n·Ts.
The following equation (1) represents the time waveform of the digital signal Xin[n·Ts].
Xin[n·Ts]=α·cos(Δθ+Σθn-1) (1)
In the equation (1), α is the amplitude factor of Xin, Δθ is the amount of change in phase during one sampling time, and Σθn-1 is an integrated value of phases at discrete-times from zero to n−1.
In the transversal filter 10, the subtraction-type first-order transversal filters 11 and 12 being the first subtraction-type first-order transversal filter D shown in
Specifically, by receiving a digital signal Xin[n·Ts], the subtraction-type first-order transversal filter 11 of the transversal filter 10 outputs, to the subtraction-type first-order transversal filter 12, a digital signal X1′[n·Ts] obtained by the following equation (2).
The subtraction-type first-order transversal filter 12 of the transversal filter 10 receives the digital signal X1′[n·Ts] from the subtraction-type first-order transversal filter 11, and outputs, to the phase frequency computer 4, a digital signal X1[n·Ts] obtained by the following equation (3).
In the transversal filter 20, the subtraction-type first-order transversal filter 21 being the first subtraction-type first-order transversal filter D shown in
Specifically, by receiving a digital signal Xin[n·Ts], the subtraction-type first-order transversal filter 21 of the transversal filter 20 outputs, to the addition-type first-order transversal filter 22, a digital signal X2′[n·Ts] obtained by the following equation (4).
The addition-type first-order transversal filter 22 of the transversal filter 20 receives the digital signal X2′[n·Ts] from the subtraction-type first-order transversal filter 21, and outputs, to the phase frequency computer 4, a digital signal X2[n·Ts] obtained by the following equation (5).
In the transversal filter 30, the addition-type first-order transversal filters 31 and 32 being the first addition-type first-order transversal filter I shown in
Specifically, by receiving a digital signal Xin[n·Ts], the addition-type first-order transversal filter 31 of the transversal filter 30 outputs, to the addition-type first-order transversal filter 32, a digital signal X3′[n·Ts] obtained by the following equation (6).
The addition-type first-order transversal filter 32 of the transversal filter 30 receives the digital signal X3′[n·Ts] from the addition-type first-order transversal filter 31, and outputs, to the phase frequency computer 4, a digital signal X3[n·Ts] obtained by the following equation (7).
The phase frequency computer 4 receives the digital signals X1[n·Ts], X2[n·Ts], and X3[n·Ts] from the transversal filters 10, 20, and 30, and computes a phase θX[n·Ts] and a frequency fX[n·Ts] of the input digital signal Xin by performing phase computation and frequency computation using the received digital signals X1[n·Ts], X2[n·Ts], and X3[n·Ts].
Specifically, by receiving the digital signals X1[n·Ts] and X2[n·Ts] from the transversal filters 10 and 20, the division computation circuitry 41 divides the digital signal X2[n·Ts] by the digital signal X1[n·Ts], as shown in the following equation (8), and outputs, to the multiplication computation circuitry 43, a division computation signal X2[n·Ts]/X1[n·Ts] being the result of the division computation.
By receiving the digital signals X2[n·Ts] and X3[n·Ts] from the transversal filters 20 and 30, the division computation circuitry 42 divides the digital signal X2[n·Ts] by the digital signal X3[n·Ts], as shown in the following equation (9), and outputs, to the multiplication computation circuitry 43, a division computation signal X2[n·Ts]/X3[n·Ts] being the result of the division computation.
The multiplication computation circuitry 43 receives the division computation signals X2[n·Ts]/X1[n·Ts] and X2[n·Ts]/X3[n·Ts] output by the division computation circuitries 41 and 42. The multiplication computation circuitry 43 multiplies the division computation signal X2[n·Ts]/X1[n·Ts] by the division computation signal X2[n·Ts]/X3[n·Ts], as shown in the following equation (10), and outputs, to the n-th root computation circuitry 44, a multiplication computation signal Y[n·Ts] being the result of the multiplication computation.
The n-th root computation circuitry 44 receives the multiplication computation signal Y[n·Ts] from the multiplication computation circuitry 43. The n-th root computation circuitry 44 performs an n-th root computation on the absolute value of the received multiplication computation signal Y[n·Ts], and outputs, to the phase computation circuitry 45, an n-th root computation signal N[n·Ts] being the result of the n-th root computation.
N[n·Ts]=tan(Δθ+Σθn-2) (11)
Here, since n for the n-th root computation is 0.5, N[n·Ts] is [|Y[n·Ts]|]0.5.
After receiving the n-th root computation signal N[n·Ts] from the n-th root computation circuitry 44, the phase computation circuitry 45 performs an arctangent computation on the n-th root computation signal N[n·Ts] according to following equation (12). The phase computation circuitry 45 further performs sign conversion on the result of the arctangent computation depending on the sign of the division computation signal X2[n·Ts]/X3[n·Ts] output by the division computation circuitry 42, and thereby computes a phase θX[n·Ts] of the digital signal Xin.
Specifically, when the sign of the division computation signal X2[n·Ts]/X3[n·Ts] is positive, the sign of the phase θX[n·Ts] of the digital signal Xin is also positive, and when the sign of the division computation signal X2[n·Ts]/X3[n·Ts] is negative, the sign of the phase θX[n·Ts] of the digital signal Xin is also negative.
The example is disclosed here, in which the phase computation circuitry 45 performs sign conversion on the result of the arctangent computation depending on the sign of the division computation signal X2[n·Ts]/X3[n·Ts] output by the division computation circuitry 42. Alternatively, the phase computation circuitry 45 may perform sign conversion on the result of the arctangent computation depending on the sign of the division computation signal X2[n·Ts]/X1[n·Ts] output by the division computation circuitry 41.
When the transversal filters 10, 20, and 30 in the filter circuitry 1 have the configurations shown in
After the phase θX[n·Ts] of the digital signal Xin is computed by the phase computation circuitry 45, the frequency computation circuitry 46 computes a frequency fX[n·Ts] of the input digital signal Xin by using the phase θX[n·Ts] of the digital signal Xin, as shown in the following equation (13).
fX[n·Ts]={θX[n·Ts]−θX[(n−1)·Ts]}·fs/(2π) (13)
In the Embodiment 1, the phase frequency computer 4 is able to obtain by the equations (12) and (13) the phase θX[n·Ts] and frequency fX[n·Ts] of the digital signal Xin without performing a Hilbert transformation.
In addition, since the number of stages of delay circuitry used by the transversal filters 10, 20, and 30 is two, the delay time resulting from detection is as short as two sampling times and thus a short-pulse input signal can be detected.
The Embodiment 1 represents that the division computation circuitry 41 computes a division computation signal X2[n·Ts]/X1[n·Ts] and the division computation circuitry 42 computes a division computation signal X2[n·Ts]/X3[n·Ts]. Alternatively, the division computation circuitry 41 may compute a division computation signal X1[n·Ts]/X2[n·Ts] and the division computation circuitry 42 may compute a division computation signal X3[n·Ts]/X2[n·Ts].
In this case, an n-th root computation signal N[n·Ts]=[|Y[n·Ts]|]0.5 output by the n-th root computation circuitry 44 is given by the following equation (14).
Here, in accordance with the mathematical formula “cot−1(X)=−tan−1(X)+π/2”, an arctangent computation is performed on a value obtained by the equation (14), and π/2 is added to the result of the arctangent computation whose sign is reversed. As a result, the phase θX[n·Ts] of the digital signal Xin has the same value as that obtained in equation (12).
As is clear from the above description, according to the Embodiment 1, it is constituted such that the filter circuitry 1 including the transversal filters 10, 20, and 30 connected in parallel to each other is provided, the transversal filters 10, 20, and 30 changing the amplitude and phase of an input digital signal Xin[n·Ts] and outputting different digital signals X1[n·Ts], X2[n·Ts], and X3[n·Ts] as respective resulting digital signals whose amplitude and phase have been changed; and the phase frequency computer 4 performs phase computation and frequency computation that use the digital signals X1[n·Ts], X2[n·Ts], and X3[n·Ts] output by the transversal filters 10, 20, and 30, and thereby computes the phase θX[n·Ts] and frequency fX[n·Ts] of the input digital signal Xin[n·Ts]. Thus, advantageous effects are achieved such that, even if a digital signal Xin[n·Ts] to be input is a short-pulse signal, a phase θX[n·Ts] and a frequency fX[n·Ts] can be detected, and increase in circuitry size and increase in power consumption, which may be caused by increase in the number of stages of delay circuitry, can be suppressed.
The Embodiment 1 represents an example, in which in the transversal filter 10 the subtraction-type first-order transversal filters 11 and 12 being the first subtraction-type first-order transversal filter D shown in
Note, however, that this is simply one example. The transversal filters 10, 20, and 30 may have any combination of types of first-order transversal filters to be used.
For example, while the second subtraction-type first-order transversal filter −D shown in
Note that the result of sign conversion performed by the phase computation circuitry 45 is reversed depending on the combination of first-order transversal filters to be used.
The above-described Embodiment 1 represents that the phase frequency computer 4 performs a multiplication computation by the multiplication computation circuitry 43 after performing a division computation by each of the division computation circuitries 41 and 42. However, the phase frequency computer 4 may perform a division computation after a multiplication computation.
Multiplication computation circuitry 71 is first multiplication computation circuitry that performs a multiplication computation between a digital signal X1[n·Ts] output by a transversal filter 10 and a digital signal X3[n·Ts] output by a transversal filter 30, and outputs a resulting signal X1[n·Ts]·X3[n·Ts] of the multiplication computation as a first multiplication computation signal.
Multiplication computation circuitry 72 is second multiplication computation circuitry that performs a square multiplication on a digital signal X2[n·Ts] output by a transversal filter 20 and outputs the resulting signal X2[n·Ts]·X2[n·Ts] of the square multiplication as a second multiplication computation signal.
Division computation circuitry 73 divides the resulting signal X2[n·Ts]·X2[n·Ts] of the square multiplication coming from the multiplication computation circuitry 72 by the resulting signal X1[n·Ts]·X3[n·Ts] of the multiplication computation coming from the multiplication computation circuitry 71. After that, the computation circuitry 73 outputs a resulting signal (X2[n·Ts]·X2[n·Ts])/(X1[n·Ts]·X3[n·Ts]) of the division computation as a division computation signal to an n-th root computation circuitry 44.
As shown in
The above-described Embodiment 1 represents that the filter circuitry 1 is constituted by the three transversal filters 10, 20, and 30, which are connected in parallel to each other. Alternatively, the filter circuitry may be constituted by two transversal filters 10 and 30, which are connected in parallel to each other.
In the Embodiment 3, a transversal filter 10 serves as a first transversal filter, and a transversal filter 30 serves as a second transversal filter.
Division computation circuitry 81 divides a digital signal X1 coming from the transversal filter 10 by a digital signal X3 coming from the transversal filter 30, and outputs, to n-th root computation circuitry 82, a division computation signal X1/X3 being the result of the division computation.
The n-th root computation circuitry 82 is root computation circuitry that performs an n-th root computation on the absolute value of the division computation signal X1/X3 output by the division computation circuitry 81, and outputs, to frequency computation circuitry 83, an n-th root computation signal N being the result of the n-th root computation.
The frequency computation circuitry 83 computes a frequency fX of an input digital signal Xin by using the n-th root computation signal N output by the n-th root computation circuitry 82.
Phase computation circuitry 84 computes a phase θX of the input digital signal Xin by using the frequency fX computed by the frequency computation circuitry 83.
Next, operation will be described.
It is assumed that a digital signal Xin[n·Ts] is input to the transversal filters 10 and 30 in the filter circuitry 1 at discrete time n·Ts in the same manner as the above-described Embodiment 1.
In the transversal filter 10, the subtraction-type first-order transversal filters 11 and 12, each of which is the first subtraction-type first-order transversal filter D shown in
Specifically, after receiving the digital signal Xin[n·Ts], the subtraction-type first-order transversal filter 11 of the transversal filter 10 outputs a digital signal X1′[n·Ts] according to the above-described equation (2) to the subtraction-type first-order transversal filter 12.
The subtraction-type first-order transversal filter 12 of the transversal filter 10 receives the digital signal X1′[n·Ts] coming from the subtraction-type first-order transversal filter 11 and outputs a digital signal X1[n·Ts] according to the above-described equation (3) to the phase frequency computer 4.
In the transversal filter 30, the addition-type first-order transversal filters 31 and 32, each of which are the first addition-type first-order transversal filter I shown in
Specifically, after receiving the digital signal Xin[n·Ts], the addition-type first-order transversal filter 31 of the transversal filter 30 outputs a digital signal X3′[n·Ts] according to the above-described equation (6) to the addition-type first-order transversal filter 32.
The addition-type first-order transversal filter 32 of the transversal filter 30 receives the digital signal X3′[n·Ts] coming from the addition-type first-order transversal filter 31 and outputs a digital signal X3[n·Ts] according to the above-described equation (7) to the phase frequency computer 4.
After receiving the digital signals X1[n·Ts] and X3[n·Ts] coming from the transversal filters 10 and 30, the phase frequency computer 4 performs phase computation and frequency computation using the digital signals X1[n·Ts] and X3[n·Ts], and thereby computes a phase θX[n·Ts] and a frequency fX[n·Ts] of the input digital signal Xin.
Specifically, by receiving the digital signals X1[n·Ts] and X3[n·Ts] from the transversal filters 10 and 30, the division computation circuitry 81 of the phase frequency computer 4 divides, as shown in the following equation (15), the digital signal X1[n·Ts] by the digital signal X3[n·Ts] and outputs to the n-th root computation circuitry 82 a division computation signal X1[n·Ts]/X3[n·Ts] being the result of the division computation.
The n-th root computation circuitry 82 receives the division computation signal X1[n·Ts]/X3[n·Ts] from the division computation circuitry 81. The n-th root computation circuitry 82 performs an n-th root computation on the absolute value of the division computation signal X1[n·Ts]/X3[n·Ts] and outputs, to the frequency computation circuitry 83, an n-th root computation signal N[n·Ts] being the result of the n-th root computation.
N[n·Ts]=tan(0.5·Δθ) (16)
Here, since n in the n-th root computation is 0.5, N[n·Ts] is [|−tan2(0.5·Δθ)|]0.5.
The frequency computation circuitry 83 receives the n-th root computation signal N[n·Ts] from the n-th root computation circuitry 82 and performs an arctangent computation on the n-th root computation signal N[n·Ts] according to the following equation (17).
Subsequently, the frequency computation circuitry 83 computes a frequency fX[n·Ts] of the digital signal Xin[n·Ts] according to the following equation (18) by using 0.5·Δθ being the result of the arctangent computation for the n-th root computation signal N[n·Ts].
Note that, according to the equation (18), the amount of change in phase Δθ for one sampling time always has a positive value as with the frequency fX[n·Ts]. Hence, phase sign conversion, such as described in the above-described Embodiment 1, is not required.
After the frequency fX[n·Ts] of the digital signal Xin[n·Ts] is computed by frequency computation circuitry 83, the phase computation circuitry 84 computes a phase θX[n·Ts] of the digital signal Xin[n·Ts], according to the following equation (19), by using the frequency fX[n·Ts] of the digital signal Xin[n·Ts].
θX[n·Ts]=Σ[fX[n·Ts]·(2π)/fs] (19)
In the Embodiment 3, by using the equations (18) and (19), the phase frequency computer 4 is able to detect the phase θX[n·Ts] and frequency fX[n·Ts] of the digital signal Xin without performing a Hilbert transformation.
In addition, since the number of stages of delay circuitry used by the transversal filters 10 and 30 is two, the delay time resulting from detection is as short as two sampling times and thus a short-pulse input signal can be detected.
The Embodiment 3 represents that the division computation circuitry 81 computes a division computation signal X1[n·Ts]/X3[n·Ts]. Alternatively, the division computation circuitry 81 may compute a division computation signal X3[n·Ts]/X1[n·Ts].
Note that, in this case, it is required to add π/2 to the result of an arctangent computation for an n-th root computation signal N[n·Ts]. By the addition of π/2, the same value as that obtained in the equation (17) is obtained.
As is clear from the above, according to the Embodiment 3, the filter circuitry 1 is constituted by the transversal filters 10 and 30 connected in parallel to each other, where the transversal filters 10 and 30 change amplitude and phase of an input digital signal Xin[n·Ts] and output different digital signals X1[n·Ts] and X3[n·Ts] as respective resulting digital signals whose amplitude and phase have been changed. Further, the phase frequency computer 4 performs phase computation and frequency computation using the digital signals X1[n·Ts] and X3[n·Ts] output by the transversal filters 10 and 30, and thereby computes the phase θX[n·Ts] and frequency fX[n·Ts] of the input digital signal Xin[n·Ts]. Thus, advantageous effects are provided that even if a digital signal Xin[n·Ts] to be input is a short-pulse signal, a phase θX[n·Ts] and a frequency fX[n·Ts] can be detected, and increase in circuitry size and increase in power consumption which result from increase in the number of stages of delay circuitry can be suppressed.
The Embodiment 3 represents an example that the transversal filter 10 is constituted by the subtraction-type first-order transversal filters 11 and 12, which are the first subtraction-type first-order transversal filter D shown in
Note, however, that this is simply one example. The transversal filters 10 and 30 may have any combination of types of first-order transversal filters to be used.
For example, while the second subtraction-type first-order transversal filter −D shown in
The above-described Embodiment 1 represents that the transversal filter 20 is constituted by the subtraction-type first-order transversal filter 21 being the first subtraction-type first-order transversal filter D shown in
In the case of
Accordingly, the computation of a phase frequency computer 4 is the same as that in the above-described Embodiment 1, and thus, the same advantageous effects as those in the above-described Embodiment 1 can be achieved.
In addition, since the subtraction-type first-order transversal filter 21 is not required for the transversal filter 20, the circuitry size can be reduced over the above-described Embodiment 1.
The above-described Embodiment 1 represents that the transversal filter 20 is constituted by the subtraction-type first-order transversal filter 21 being the first subtraction-type first-order transversal filter D shown in
In the case of
The connection order of the addition-type first-order transversal filter 31 and the subtraction-type first-order transversal filter 21 composing the above-described transversal filter is opposite to that of the subtraction-type first-order transversal filter 21 and the addition-type first-order transversal filter 22 in the transversal filter 20 of
Hence, a digital signal X2[n·Ts] to be output by the transversal filter 20 of
Therefore, the computation of a phase frequency computer 4 is the same as that in the above-described Embodiment 1, and thus, the same advantageous effects as those in the above-described Embodiment 1 can be obtained.
In addition, since the transversal filter 20 does not require the addition-type first-order transversal filter 22, the circuitry size can be reduced over the above-described Embodiment 1.
The above-described Embodiment 1 represents that an input-side first-order transversal filter in the transversal filter 20 is the subtraction-type first-order transversal filter 21, an output-side first-order transversal filter is the addition-type first-order transversal filter 22, and the transversal filter 30 is constituted by the addition-type first-order transversal filters 31 and 32, each of which is the first addition-type first-order transversal filter I shown in
In the Embodiment 6, as shown in
In addition, a transversal filter 30 is constituted by an addition-type first-order transversal filter 32 to which is input a digital signal output by the addition-type first-order transversal filter 22 in the transversal filter 20. The addition-type first-order transversal filter 32 is the first addition-type first-order transversal filter I shown in
In the case of
In addition, a digital signal X2′[n·Ts] to be output by the addition-type first-order transversal filter 22 in the transversal filter 20 indicates the same value as a digital signal X3′[n·Ts] to be output by the addition-type first-order transversal filter 31 in the transversal filter 30 of
Hence, the computation of a phase frequency computer 4 is the same as that in the above-described Embodiment 1, and thus, the same advantageous effects as those in the above-described Embodiment 1 can be obtained.
In addition, since the transversal filter 30 does not require the addition-type first-order transversal filter 31, the circuitry size can be reduced over the above-described Embodiment 1.
The above-described Embodiment 1 represents that the transversal filter 10 is constituted by the subtraction-type first-order transversal filters 11 and 12, each of which is the first subtraction-type first-order transversal filter D shown in
In
One-sampling time delaying circuitry 92 is first delaying circuitry that delays one digital signal split by the splitter circuitry 91 by one sampling time Ts.
One-sampling time delaying circuitry 93 is second delaying circuitry that delays, by one sampling time Ts, the digital signal which has been delayed by the one-sampling time delaying circuitry 92.
Coefficient multiplication circuitry 94 multiplies, by a factor of −2, the digital signal having been delayed by the one-sampling time delaying circuitry 92.
Addition computation circuitry 95 performs an addition computation among the digital signal delayed by the one-sampling time delaying circuitry 93, the digital signal multiplied by the factor by the coefficient multiplication circuitry 94, and the other digital signal split by the splitter circuitry 91.
Next, operation will be described.
The addition computation circuitry 95 adds together: a digital signal delayed by two sampling times Ts through the one-sampling time delaying circuitries 92 and 93; a digital signal that has been delayed by one sampling time Ts through the one-sampling time delaying circuitry 92 before being multiplied by a factor of −2 at the coefficient multiplication circuitry 94; and a digital signal split by the splitter circuitry 91. A transfer characteristic DF1 of the transversal filter 10, which has the constitution shown in
DF1=Z−2−2·Z−1+1 (20)
In equation (20), Z−1 is a complex number of exp(−jω·Ts) and denotes a delay of one sampling time. Z−2 represents a delay of two sampling times.
On the other hand, the transfer characteristic DF1 of the transversal filter 10 shown in
DF1=(1−Z−1)(1−Z−1)=Z−2−2·Z−1+1 (21)
According to the equations (20) and (21), the transfer characteristic DF1 of the transversal filter 10 of
The above-described Embodiment 1 represents that the transversal filter 20 is constituted such that the subtraction-type first-order transversal filter 21 being the first subtraction-type first-order transversal filter D shown in
In
One-sampling time delaying circuitry 102 is first delaying circuitry that delays one digital signal split by the splitter circuitry 101 by one sampling time Ts.
One-sampling time delaying circuitry 103 is a second delaying circuitry that delays, by one sampling time Ts, the digital signal which has been delayed by the one-sampling time delaying circuitry 102.
Coefficient multiplication circuitry 104 multiplies, by a factor of −1, the digital signal having been delayed by the one-sampling time delaying circuitry 103.
Addition computation circuitry 105 performs an addition computation between the digital signal multiplied by the factor by the coefficient multiplication circuitry 104 and the other digital signal split by the splitter circuitry 101.
Next, operation will be described.
The addition computation circuitry 105 adds together: a digital signal having been delayed by two sampling times Ts through the one-sampling time delaying circuitries 102 and 103 before being multiplied by a factor of −1 by the coefficient multiplication circuitry 104; and a digital signal split by the splitter circuitry 101. The transfer characteristic DF2 of the transversal filter 20 having the constitution shown in
DF2=1−Z−2 (22)
On the other hand, the transfer characteristic DF2 of the transversal filter 20 of
the subtraction-type first-order transversal filters 21 which is the first subtraction-type first-order transversal filter D shown in
DF2=(1−Z−1)·(1+Z−1)=1−Z−2 (23)
According to the equations (22) and (23), the transfer characteristic DF2 of the transversal filter 20 of
The above-described Embodiment 1 represents that the transversal filter 30 is constituted by the addition-type first-order transversal filters 31 and 32, each of which is the first addition-type first-order transversal filter I shown in
In
One-sampling time delaying circuitry 112 is first delaying circuitry that delays one digital signal split by the splitter circuitry 111 by one sampling time Ts.
One-sampling time delaying circuitry 113 is second delaying circuitry that delays, by one sampling time Ts, the digital signal having been delayed by the one-sampling time delaying circuitry 112.
Coefficient multiplication circuitry 114 multiplies, by a factor of 2, the digital signal having been delayed by the one-sampling time delaying circuitry 112.
Addition computation circuitry 115 is circuitry that performs an addition computation among the digital signal delayed by the one-sampling time delaying circuitry 113, the digital signal multiplied by the factor by the coefficient multiplication circuitry 114, and the other digital signal split by the splitter circuitry 111.
Next, operation will be described.
The addition computation circuitry 115 adds together: a digital signal delayed by two sampling times Ts through the one-sampling time delaying circuitries 112 and 113; a digital signal delayed by one sampling time Ts by the one-sampling time delaying circuitry 112 before being multiplied by a factor of 2 at the coefficient multiplication circuitry 114; and a digital signal split by the splitter circuitry 111. The transfer characteristic DF3 of the transversal filter 30 shown in
DF3=Z−2+2·Z−1+1 (24)
On the other hand, the transfer characteristic DF3 of the transversal filter 30 shown in
DF3=(Z−1+1)·(Z−1+1)=Z−2+2·Z−1+1 (25)
According to the equations (24) and (25), the transfer characteristic DF3 of the transversal filter 30 shown in
The above-described Embodiment 1 represents the filter circuitry 1 that is constituted by the transversal filters 10, 20, and 30. Each of the transversal filters 10, 20, and 30 may be constituted to include a digital filter for removing noise which has been superimposed on an input digital signal Xin.
A digital filter 13 removes noise superimposed on a digital signal which is output by a subtraction-type first-order transversal filter 12.
A digital filter 23 removes noise superimposed on a digital signal which is output by an addition-type first-order transversal filter 22.
A digital filter 33 removes noise superimposed on a digital signal which is output by an addition-type first-order transversal filter 32.
The circuitry configuration of the digital filters 13, 23, and 33 is not particularly limited as long as the ones are capable of removing noise superimposed on a digital signal Xin. For example, an FIR (Finite Impulse Response) filter or an IIR (Infinite Impulse Response) filter may be used as each digital filter.
Note that, in order to detect a phase θX[n·Ts] and a frequency fX[n·Ts] of a digital signal Xin[n·Ts] by the phase frequency computer 4 of
The Embodiment 10 represents that each of the digital filters 13, 23, and 33 of the corresponding transversal filters 10, 20, and 30 is disposed as an output stage of a series of the two first-order transversal filters. Alternatively, as shown in
Further alternatively, as shown in
A phase frequency computer 121 has the same configuration as the phase frequency computer 4 of
A phase frequency computer 122 has the same configuration as the phase frequency computer 4 of
As presented in the above-described Embodiments 1 and 3, the phase frequency computer 4 of
A statistics computer 123 performs a statistical computation using the phase θX computed by the phase frequency computer 121 and the phase θX computed by the phase frequency computer 122, and also performs a statistical computation using the frequency fX computed by the phase frequency computer 121 and the frequency fX computed by the phase frequency computer 122.
Next, operation will be described.
The phase frequency computer 121 computes a phase θX[n·Ts] and a frequency fX[n·Ts] of a digital signal Xin[n·Ts] in the same manner as the phase frequency computer 4 of
The phase frequency computer 122 computes a phase θX[n·Ts] and a frequency fX[n·Ts] of the digital signal Xin[n·Ts] in the same manner as the phase frequency computer 4 of
If the number of bits applied to digital computation performed by the phase frequency computers 121 and 122 is infinite, a quantization error does not occur in the digital computation. In this case, a phase θX[n·Ts] obtained by the equation (12) and a phase θX[n·Ts] obtained by the equation (19) have the same value as each other, and a frequency fX[n·Ts] obtained by the equation (13) and a frequency fX[n·Ts] obtained by the equation (18) have the same value as each other.
However, since the number of bits applied to digital computation performed by the phase frequency computers 121 and 122 has a finite value, the quantization error may occur.
Hence, the statistics computer 123 performs a statistical computation on the phase θX[n·Ts] computed by the phase frequency computer 121 and the phase θX[n·Ts] computed by the phase frequency computer 122, and also performs a statistical computation on the frequency fX[n·Ts] computed by the phase frequency computer 121 and the frequency fX[n·Ts] computed by the phase frequency computer 122.
Possible statistical computations for the two phases θX[n·Ts] computed by the phase frequency computers 121 and 122 may be a computation of finding an average value of the two phases θX[n·Ts], a computation of finding a weighted addition value, etc.
In addition, possible statistical computations for the two frequencies fX[n·Ts] computed by the phase frequency computers 121 and 122 may be a computation of finding an average value of the two frequencies fX[n·Ts], a computation of finding a weighted addition value, etc.
By those computations, the influence of quantization error occurring in digital computation of the phase frequency computers 121 and 122 can be reduced, enabling to obtain a highly accurate phase θX[n·Ts] and frequency fX[n·Ts].
The Embodiment 11 represents that the phase frequency computers 121 and 122 having different processing contents of phase computation and frequency computation are mounted, and the statistics computer 123 performs a statistical computation on the computation results of the phase frequency computers 121 and 122. Alternatively, three or more phase frequency computers having different processing contents of phase computation and frequency computation may be mounted, and the statistics computer 123 may perform a statistical computation on the computation results of the three or more phase frequency computers.
The Embodiment 11 represents that the phase frequency computers 121 and 122 having different processing contents of phase computation and frequency computation are mounted. Alternatively, filter circuitries 1 connected to previous stages to a plurality of phase frequency computers may have different configurations, and the statistics computer 123 may perform a statistical computation on the computation results of the plurality of phase frequency computers.
Specifically, for example as shown in
Also in this case, the influence of quantization error occurring in digital computation of the two phase frequency computers 121 is reduced, enabling to obtain a highly accurate phase θX[n·Ts] and frequency fX[n·Ts].
Note that free combinations of the embodiments, or modifications to any component in the embodiments, or omissions of any component in the embodiments which fall within the spirit and scope of the invention may be made to the invention of the present application.
Phase frequency detection devices according to the invention are suitable for devices that need to detect a phase θX and a frequency fX of an input digital signal Xin even if the input digital signal Xin is a short-pulse signal.
1: filter circuitry, 4: phase frequency computer, 10: transversal filter (first transversal filter), 11 and 12: subtraction-type first-order transversal filter, 13: digital filter, 20: transversal filter (second transversal filter), 21: subtraction-type first-order transversal filter, 22: addition-type first-order transversal filter, 23: digital filter, 30: transversal filter (third transversal filter, second transversal filter), 31 and 32: addition-type first-order transversal filter, 33: digital filter, 41: division computation circuitry (first division computation circuitry), 42: division computation circuitry (second division computation circuitry), 43: multiplication computation circuitry, 44: n-th root computation circuitry (root computation circuitry), 45: phase computation circuitry, 46: frequency computation circuitry, 51: splitter circuitry (first splitter circuitry), 52: one-sampling time delaying circuitry (first delaying circuitry), 53 and 54: subtraction computation circuitry, 61: splitter circuitry (second splitter circuitry), 62: one-sampling time delaying circuitry (second delaying circuitry), 63 and 64: addition computation circuitry, 71: multiplication computation circuitry (first multiplication computation circuitry), 72: multiplication computation circuitry (second multiplication computation circuitry), 73: division computation circuitry, 81: division computation circuitry, 82: n-th root computation circuitry (root computation circuitry), 83: frequency computation circuitry, 84: phase computation circuitry, 91: splitter circuitry, 92: one-sampling time delaying circuitry (first delaying circuitry), 93: one-sampling time delaying circuitry (second delaying circuitry), 94: coefficient multiplication circuitry, 95: addition computation circuitry, 101: splitter circuitry, 102: one-sampling time delaying circuitry (first delaying circuitry), 103: one-sampling time delaying circuitry (second delaying circuitry), 104: coefficient multiplication circuitry, 105: addition computation circuitry, 111: splitter circuitry, 112: one-sampling time delaying circuitry (first delaying circuitry), 113: one-sampling time delaying circuitry (second delaying circuitry), 114: coefficient multiplication circuitry, 115: addition computation circuitry, 121 and 122: phase frequency computer, and 123: statistics computer
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/JP2015/056365 | 3/4/2015 | WO | 00 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2016/139778 | 9/9/2016 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
5444416 | Ishikawa et al. | Aug 1995 | A |
20030052662 | Bi et al. | Mar 2003 | A1 |
Number | Date | Country |
---|---|---|
S63-258105 | Oct 1988 | JP |
H06-216651 | Aug 1994 | JP |
H106-247245 | Sep 1994 | JP |
H06-291553 | Oct 1994 | JP |
2000-341348 | Dec 2000 | JP |
2005-091255 | Apr 2005 | JP |
Entry |
---|
Allen, Jr. et al., A Unified Approach to Transform Domain LMS Adaptive Filtering, 1988 IEEE, pp. 424-427. |
International Search Report issued in PCT/JP2015/056365; dated May 12, 2015. |
Written Opinion issued in PCT/JP2015/058365; dated May 12, 2015. |
Number | Date | Country | |
---|---|---|---|
20180003750 A1 | Jan 2018 | US |