JPS5838698Y2 - Multi-frequency signal receiving device - Google Patents

Description

[Detailed Description of the Invention] The present invention provides a multi-frequency signal receiving device that conveys information by a combination of a plurality of predetermined signal frequencies;
In particular, a multi-frequency interoffice selection signal (hereinafter referred to as an MF signal) used in telephone switching systems.

) The present invention relates to a receiving device suitable for receiving.

Specific examples of multi-frequency signals include: (1) MF signal: 700.900.1100.130
Information is conveyed by a combination of two arbitrary frequency signals of six waves of 0.1500.1700 Hz, and the types of information that can be conveyed are as follows.

C2-6×5/2=15.

(2) Push button signal (hereinafter referred to as PB double signal).

): Low frequency 4 waves (697,770,852,941Hz)
Information is conveyed by a combination of any one frequency within each band, including three high-frequency waves (1209, 1336° and 1477 Hz), and the number of types of information that can be conveyed is 4×3 = 12 types.

Conventionally, in order to detect these multi-frequency signals, detection methods using a group of band-pass filters made in advance according to the frequency (analog/digital filter method), detection by correlation with sine and cosine signals corresponding to the frequency are used. Law (
Various methods have been proposed, including analog/digital spectrum analysis (analog/digital spectrum analysis), period measurement by measuring the zero-crossing interval of the waveform, but these methods involve unnecessary calculations even for frequency bands with no input signal components. normal sampling frequency for multi-frequency signals, e.g. 4K
There were problems such as not being able to obtain sufficient accuracy at Hz.

Therefore, in the former case, the number of parts and time required for processing increases, and the cost of the receiver increases.In the latter case, the multiprocessing ability of the receiver decreases, and the receiver The disadvantage was that it led to an increase in costs.

In order to eliminate such drawbacks, the present inventors previously proposed a receiving system as disclosed in Japanese Patent Application No. 110590/1982.

This reception method uses PAR, which is one of the linear predictive analysis techniques.
COR analysis method (for example, shown in Japanese Patent Publication No. 49/18007).

) is applied to analyze the received PB signal, and the partial autocorrelation coefficient (PARCOR coefficient, also called k parameter) is applied.

), the number of parameters required to identify the waveform of a two-frequency sine wave signal, basically four, are extracted, and the receiving frequency is identified based on this. Compared to the conventional method, Because the reception frequency can be determined from the processing of the reception signal itself, and because it can be processed sufficiently at a low sampling frequency, for example, 4KHz, it is possible to receive with high precision and high multiplicity, and it is also possible to perform processing with a microcomputer, etc. This makes it possible to reduce the cost of the receiver, and it is also resistant to so-called pseudo-signaling, where audio waveforms other than multi-frequency signals are mistakenly received as signals. .

However, such a receiving method based on the PARCOR analysis method actually contains a 400H2 signal in the received signal, which indicates whether the phone is available or busy when the handset is picked up. There is a drawback that it is necessary to provide a high-pass filter or a bandpass filter to remove the noise.

In addition, in the above reception method, even considering the condition that in principle only the quartic equation, the conjugate complex symmetry of the root, and the frequency need to be determined in order to determine the signal frequency of the two frequencies,
Since it is necessary to solve quadratic equations, signal processing becomes complicated and processing time becomes longer, and if a circuit is provided to solve quadratic equations, the circuit becomes complicated and expensive. .

Therefore, the present inventors first applied the above-mentioned patent application No. 521105.
As an improvement to No. 90, we focused on the fact that in multi-frequency signals such as PB signals, the signal frequency band is limited to specific low and high frequency bands, and each band contains only one frequency. Then, the frequency spectrum of the received signal is analyzed, and a spectral part of a specific band containing only one frequency signal is selectively extracted from the spectrum, and a partial autocorrelation coefficient is determined from the power spectrum within that band. proposed a multi-frequency signal reception method using a simplified method that allows the signal frequencies included in the band to be determined by solving a linear equation, thereby shortening the processing time and reducing the cost of the equipment. ing.

However, such a reception method can be applied to reception of a PB signal in which one frequency is included in each of the specific low and high bands, but it cannot be applied to reception of an MF multiplied signal in which two frequencies are included in any two bands. If the reception method of the simplified method, which has the effectiveness described above, is applied to the reception of the PB signal, it is necessary to process the MF multiplied signal using a different reception method than the PB signal, so both signals are There was a problem in that it was not possible to standardize the receiving device, and the cost of the device was high.

The present invention arbitrarily limits the analysis frequency band of a received multi-frequency signal, calculates the ratio of the power within the limited range to the total power, and determines the frequency band in which the multi-frequency signal is included based on the ratio. By determining the partial autocorrelation coefficient within that band and identifying the multifrequency signal, the MF multiplied signal P
The present invention provides a multi-frequency signal receiving device that can perform integrated processing together with the B signal.

First, the principle of the present invention will be explained.

MF double signal, as is well known, six sine wave signals are assigned at 200 Hz intervals from 700H2 to 1700 Hz, and any combination of two frequencies among these is used as a signal. .

As mentioned above, the total number is 15 types.

Now, with 1200H2 as the border, 700-1100Hz
Low range between 600 and 1200 Hz, including 1300 and 1200 Hz
If we define the range between 1200 and 1800 Hz, including 1700H2, as the high range, if there is one wave of each signal in this low and high frequency range, there are 3x3=9 types, and 315 out of 15 types, i.e. It accounts for the majority.

The 6 types other than these 9 types are 3 types that are in the low range for both two frequencies.
Types (700Hz and 900Hz, 700Hz and 1100
Hz, 900 Hz and 1100Hz), three types with both frequencies in the high range (1700Hz and 1500Hz, 17
00Hz and 1300Hz, 1500Hz and 1300
Hz).

Each of these three types has a frequency of 600Hz, including 700H2 to 900Hz.
When analyzing 1400 to 1800 Hz including ~1000 Hz or 1500 H2 to 1700 H2 as one band, if one frequency f and/or f H' exists there, it is a combination of that and 1100 Hz or 1300 Hz, and there If there are two frequencies, it is a combination of 700 Hz and 900'Hz or 1500 Hz and 1700 Hz.

In the case of the MF multiplied signal, unlike the case of the PB signal, the relative level difference between the two frequency signals to be detected and received is 7 dB.
This is smaller than that for the B signal (15 dB).

Therefore, the received two-frequency signal has one frequency in the low frequency range and one in the high frequency range.
When each wave can be separated, the ratio ρ of the in-band power to the total power will be between 0.3 and 0.7.

That is, the power ratio of low and high frequencies is 7 dB (=2.
25), and the total power is 2.25+1=3.25
Since it is expressed as
corresponds to 0.3 to 0.7.

). Furthermore, if the power ratio ρ becomes larger than 0.7 or smaller than 0.3, the two-frequency signal will coexist in either the low frequency band or the high frequency band.

FIG. 1 is a flowchart showing an example of an MF signal detection accompaniment logic according to the present invention, the details of which will be explained based on an embodiment of the multi-frequency signal receiving device shown in FIG.

In addition, in Fig. 1, the process starts from checking the ratio ρ1 of the low-range power to the total power, but the same processing flow can be obtained even if it starts from the ratio pH of the high-range power to the total power. .

Also, when two frequencies are considered to coexist in the low or high range, only one type of band division is performed;
00 Hz and 900 to 1100 Hz) and check the results in combination. In that case, more accurate detection will be possible.

As mentioned above, FIG. 2 shows the configuration of an embodiment of the multi-frequency signal receiving device according to the present invention.

A multi-frequency signal waveform (assuming it is digitized) whose presence is to be analyzed and detected.

If it is an analog waveform, it is necessary to perform A/D conversion and digitize it.

) 100 is determined by switch 1, for example lQms
The data are stored alternately in the two buffer memories 2 and 3.

This is because data reading and processing are performed in parallel, and processing is performed at the time of execution.

Therefore, for example, while data is being read into the buffer memory 2, signal detection processing is performed on data that has already been read into the buffer memory 3.

Switch 1 switches from buffer memory 3 side to buffer memory 2
In synchronization with the switching to the buffer memory 3 side, the switch 4 is switched from the buffer memory 2 side to the buffer memory 3 side, and processing for the data read into the buffer memory 3 is started.

First, a fast Fourier transform (FFT) processor 5 converts the received waveform into its power spectrum.

The signal existence range of the converted power spectrum (e.g. 6
The received signal total power Vo is determined as the sum of the components (00 to 1800 Hz) and is stored in the buffer memory 6.

At the same time, the low range (e.g. 600-1200H
z) and high frequency (for example, 1200 to 1800 Hz) energy, that is, the power spectrum, is determined and stored in buffer memories 7 and 8.

By the way, in the case of MF signals, unlike PB signals, there is no audio signal, so
In a section where the signal energy between Hz shows a signal level equal to or higher than the level defined as the signal level to be received, it may be treated as if a signal exists.

Therefore, the energy of the entire signal ■.

■ There are preset thresholds.

(Stored in buffer memory 9.

) is compared by the comparison circuit 10, and if V o > Vo, the signal is determined to be present and detected.

If Vo<Vo, it is determined that it is not a signal section and the process is stopped.

The result is output as signal 101.

If vo>vo, the arithmetic circuit 11 calculates cosin from the low-frequency power spectrum in the buffer memory 7.
Low-frequency autocorrelation coefficient ■ by e-transform.

1-9V1[,...,V2t are determined and recorded in the buffer memory 12 (the subscript L indicates the low frequency range, and 0 and 1.2 indicate the delay time.

). Next, in the arithmetic and comparison circuit 13, Vo+- is
.

By normalizing with ρ. 1. =Vo+-/Vo
(low frequency power ratio) is determined, and the control switch 14 is operated depending on the magnitude thereof.

As shown in the flowchart in Figure 1, for example ゛, 0
．． 3〈ρ.

When 1.degree.

Also, ρ.

, > t17, the signal 103 is output and ρo, -
When <o, 3, a signal 104 is output and processing to be described later is performed.

When 2L is determined for the above-mentioned partial autocorrelation coefficient kLL, the value of k2L is compared in the comparator circuit 16.

Generally, if only one frequency sine wave is included in a given band, the value of the second-order partial autocorrelation coefficient determined from the power spectrum of that band will be approximately -1, so
Check the second-order partial autocorrelation coefficient and find that its value is -0,
If it is smaller than 8, it can be determined that a sine wave of one frequency substantially exists within that band.

Therefore, if the comparison circuit 16 checks that the partial autocorrelation coefficient is smaller than the threshold -0.8, then
Since one sine wave signal exists in the low frequency range, the comparison circuit 17 is activated by the signal 105 at that time.

In this comparison circuit 17, the first-order partial autocorrelation coefficient determined by the arithmetic circuit 15 is 1□7, and the three signal frequencies stored in the buffer memory 18 are 700 Hz and 900 Hz.
The partial autocorrelation coefficient corresponding to each Hz.1100 car is compared with a value of 1, the signal frequency f is determined, and the signal 106 is sent to the decoding circuit 19.

During this time, if the coefficient 21 is larger than -0.8, and if the value of the coefficient klL is not within the pre-stored value range of 1, this is a logical error or another error signal. and reject signals 107 and 10 as
Outputs 8.

The comparator circuit 16 checks the coefficient for a value of 21-, and if it is smaller than -0,8, the signal of the other wave is in the high range, so the comparator circuit 16 sends the signal 109 to the arithmetic circuit 20. , the arithmetic circuit 20 performs Co51ne transformation from the high-frequency power spectrum of the buffer memory 8 to obtain a high-frequency autocorrelation coefficient V.

11° VIH and V2H are determined and stored in the buffer memory 21 (the subscript H represents the high frequency range, and 0 and 1.2 represent the delay time).

). From this, the arithmetic circuit 22 calculates 21□ for the first-order and second-order partial autocorrelation coefficients kLH, and the comparison circuit 23 checks that kzll<0.8. If k2n<0.8, the signal 110 is output to the comparison circuit 24 and stored in the buffer memory 25.
``Three signal frequencies 1300, 1500,
The comparison circuit 24 compares the coefficients corresponding to 1700 Hz with the value of 1 and the coefficients obtained by the arithmetic circuit 22 to determine the signal frequency fH, and sends it to the decoding circuit 19.

During this time, when kzn > 0.8 and the coefficient klH
If it is not within the range of 1 in the pre-stored coefficients, there is a logical inconsistency or the signal frequency is incorrect, and the signals are rejected and signals 111 and 112 are output.

The decoding circuit 19 decodes the signal information from the value of the low signal frequency f1- and the value of the high signal frequency f1□, and outputs the result as a signal 113.

Now, the judgment condition for the switch circuit 14 is ρt, > 0.7
, that is, when the signal 103 is output, 2
This shows that both signal frequencies exist in the low range.

The processing at that time will be explained next. As mentioned above, FFT
The processor 5 calculates the spectrum (power spectrum) of the signal waveform.

The signal 103 from the switch circuit 14 changes the power spectrum of the processor 5 from 600 to 1000, for example.
The Hz range is extracted and stored in the buffer memory 26.

The arithmetic circuit 27 performs Co51ne conversion to obtain the correlation coefficient ■.

1/. Vo, /, y 2+-/ is determined and stored in the buffer memory 28 .

From this correlation coefficient, the calculation circuit 29 calculates the following partial autocorrelation coefficient and stores it in the buffer memory 30.

Next, the comparison circuit 31 checks that 2+-'<0.8, and k2. When '< o, s, the comparison circuit 33 compares the coefficient klL' of the buffer memory 30 with the value of 1 to the coefficients corresponding to the frequencies of 700 Hz and 900I-1z stored in advance in the buffer memory 32, Determine its frequency f, -/.

At this time, since it is logically determined that the other wave is 1100 Hz, the decoding circuit 34 decodes the signal as 1100 Hz and f14', and outputs the result as the signal 114.

Also, as a result of the check in the comparison circuit 31, k2. '<0.
8, two signal frequencies will coexist in this band: 700 Hz and 900 Hz.

The result is output as an output signal 115. Also, the values of the coefficients k l+,,'
If the value of the coefficient does not match within the range of 1, the signal is an error, and the result is output as the signal 116.

Furthermore, the judgment condition of the switch circuit 14 is ρr-<0
, 3, that is, when the signal 104 is output, it means that the two signal frequencies both exist in the high range.

The processing at that time will be explained next. A spectrum in the range of 1400 to 1800 Hz, for example, is extracted from the power spectrum of the processor 5 using the signal 104 from the switch circuit 14 and stored in the buffer memory 35 .

The spectrum in the buffer memory 35 is cosine-transformed by the arithmetic circuit 36 to obtain the correlation coefficient (■).

H', V u-+', and V2H' are determined and stored in the buffer memory 37.

From this correlation coefficient, the calculation circuit 38 calculates the following partial autocorrelation coefficient and stores it in the buffer memory 39.

Next, the comparator circuit 40 calculates kz+q'<0.8
When k2n'< 0.8, the comparison circuit sets 111' to the coefficient of the buffer memory 39 and sets the value of 1 to the coefficient corresponding to the frequencies of 1500 Hz and 1700 Hz stored in advance in the buffer memory 41. 42 to determine the frequency fH'.

At this time, it is logically determined that the other frequency is 1300Hz, so the decoding circuit 43 selects 1300Hz.
The signal is decoded as fH' and the result is output as signal 117.

Also, as a result of the check in the comparator circuit 40, k2n'<
0.8, two signal frequencies will coexist in this band: 1500 Hz and 170 Hz.
0 Hz.

The result is output as signal 118.

Also, the value of coefficient klH' is 1500 Hz and 1700H
If the coefficient of z is not within the range of 1, the signal is erroneous;
The result is output as signal 119.

As described above, according to the present invention, the reception of a multi-frequency signal such as an MF signal that includes one frequency in any two bands out of a large number of bands can be performed in the low frequency range such as a PB signal. It is possible to integrally receive multi-frequency signals including one frequency in each of the high and high frequencies, and a very inexpensive receiving device can be obtained.

[Brief explanation of drawings]

FIG. 1 is a flowchart showing an example of a receiving method according to the present invention, and FIG. 2 is a block diagram showing an embodiment of a receiving apparatus according to the present invention. 1.4...Switch, 5...FFT processor, 6-9...Buffer memory, 10,
16.17... Comparison circuit, 11.15...
...Arithmetic circuit, 19...Decoding circuit.

Claims (1)

[Scope of utility model registration request] A first means for determining the power spectrum of a received multi-frequency signal, and a ratio of the power in an arbitrary limited frequency band to the power in all frequency bands in the multi-frequency signal, using the information of the first means. Find, and by the ratio,
second means for determining a frequency band to be analyzed;
The power spectrum of the multi-frequency signal within the frequency band determined by the means is extracted from the first means, the partial autocorrelation coefficient of the multi-frequency signal within the frequency band is determined, and the frequency of the multi-frequency signal is determined by A multi-frequency signal receiving device comprising: third identifying means.