JPH0225317B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an audible signal identification circuit for an audible signal generation circuit in a time division digital exchange by periodically reading out a digital code sequence stored in a memory.

Time division digital switching equipment is becoming the most popular switching equipment due to rapid advances in digital communication technology and LSI technology. Furthermore, due to its affinity for digital signals, its scope of application is expanding not only to telephones but also to data terminals and data processing devices. Especially from the point of view of streamlining business communications,
The situation calls for the emergence of integrated voice/data switches that accommodate both telephone and data equipment and exchange them all at once.

There are various ways to implement such an integrated voice/data switch. For example, it is possible to use a packet switching method to provide the exchange itself with advanced functions, but it is also possible to achieve this by interconnecting devices with equivalent functions using a circuit switching method. This is because, in this method, the exchange itself becomes extremely simple, and the restrictions on connections with terminals and the like are relaxed.

There are several possible circuit switching methods. The big issue here is how to deal with differences in terminal types. Particularly problematic is the difference in line control signals between the terminal and the exchange, such as line setting, disconnection, and transmission of selection signals. On the other hand, the method of using different line control signals for each terminal has the disadvantage that the control program for the exchange becomes complicated, and the type of terminal must be registered, which causes inconvenience when changing. On the other hand, if an interface circuit is provided at the input section from the terminal to the exchange, and this interface circuit unifies the line control signals so that all terminals look exactly the same from the control section of the exchange, the above-mentioned method can be achieved. It has no problems and is extremely advantageous. Furthermore, since telephones are considered to be the most common terminal in business communications, a method of unifying the line control signals of each terminal into the line control signal of the telephone is promising.

In other words, in the above-mentioned interface circuit, the signals STX, STX, and
Converts ETX to off-hook and on-hook signals,
Converts a selection signal from a character dial or the like into a dial pulse or push button signal.
On the other hand, the exchange says that it is ready to receive dialing, and that it is busy.
During a call, an audible signal (such as a dial tone, busy tone, ringback tone, etc.) is sent to indicate the progress of the call. The interface circuit must convert these signals into ACK, NACK, Ready, Wait, etc. signals for data terminals and the like. Therefore, the interface circuit requires an audible signal identification circuit.

However, since these audible signals have conventionally been used to prompt humans to perform actions and to determine the progress of a call, it can be said that there are almost no examples of circuits in which these signals can be recognized by machines. However, by analogy with the function of identifying a specific frequency from among audio signals, if an audible signal identifying circuit is configured in the manner of a push-button signal receiver (PBR), it will be as shown in FIG. 1. Note that these audible signals are, in principle, signals expressed in the form of the sum of sine waves of two frequencies.

In FIG. 1, since the digital code word applied to the input terminal 10 is normally a nonlinear code that has been subjected to companding, it is converted into a linear code by the decompression circuit 11, and then passed through the low-pass filter 12 and the high-frequency code word. It is added to the filter 13. Both of these filters are digital filters, and the cutoff frequency is determined depending on the two frequencies contained in the audible signal to be discriminated. For example, in the case of a dial tone consisting of two frequencies, 350Hz and 440Hz,
Simply put, the boundary is 395Hz, which is the intermediate frequency between the two, and the frequency components below that are filtered through band filter 1.
4, the higher frequency components are passed through a band filter 15.
Add to. The filters 14 and 15 are narrow band filters whose center frequencies are two frequencies included in the audible signal to be identified, and are also constructed from digital filters. In the dial tone example described above, filter 14 is a band filter with a center frequency of 350 Hz and filter 15 is a band filter with a center frequency of 440 Hz. In this way, the desired frequency components are extracted using the filter and added to the detection circuits 16 and 17 to measure the level of each frequency component. The identification logic circuit 18 identifies the presence or absence of an audible signal from the measured level, and only when an audible signal is present, outputs the type of the audible signal to an output terminal 19.
Output to.

If there are several types of audible signals, a set of low-pass filters, high-pass filters, and bandpass filters 20, 21, 22, 23 or 26, 27, 2 is provided for each type of audible signal.
8, 29 and detection circuits 24, 25, 30, 31, etc. are provided to identify the presence or absence of each audible signal. Dial·
Examples of audible signals other than tones include ringback tones: 440Hz and 480Hz, busy tones:
There are 480Hz and 620Hz, etc.

In the conventional audible signal identification circuit as described above, the frequency components are 350Hz and 440Hz in the above example.
There are four types: Hz, 480Hz, and 620Hz, but the combined frequencies and frequency differences are different, and the smallest frequency difference is 40Hz, which is quite small. Therefore, the low-pass filters 12, 20,
26. The cutoff frequencies of the high-pass filters 13, 21, and 27 are different, and a high-order filter is required to deal with small frequency differences. This increases the number of filters and the complexity of the individual filters.

In addition, similar to the case of PBR, such audible signal identification circuits have various characteristic standards set in order to prevent malfunctions with respect to audio signals and to quickly and accurately identify desired audible signals. The order, coefficients, etc. of the filter must be optimally designed to meet the standards. A further problem is that
The standard for audible signals such as dial tones and busy tones is not as unified and standardized as that for push button signals (PB signals). In other words, it is necessary to optimally design filters and the like to satisfy the characteristic standards as described above for various audible signals each time, which is complicated and inefficient, and increases the cost due to the increased amount of design work. It also leads to an increase. Furthermore, it is not desirable to include such a complicated circuit in each interface circuit provided for each terminal. Therefore, the method of unifying line control signals using an interface circuit is also less feasible.

Although the above description has been made assuming a digital filter, it is of course possible to convert a digital code into an analog signal by a decoding circuit and construct an audible signal identification circuit using an analog filter. However, even in that case, the design of the filter will be more complicated and difficult than in the digital case.

The present invention eliminates such shortcomings in the prior art, and also takes advantage of the fact that the audible signal is generated by periodic reading of a digital code sequence stored in a memory. Another object of the present invention is to obtain an audible signal identification circuit with high identification accuracy.

That is, according to the present invention, in an audible signal identification circuit of a digital exchange having an audible signal generation circuit that generates an audible signal by periodically reading out a digital code sequence stored in a memory,
A digital code word is used as an address, and whether or not the digital code word exists in a digital code series representing each audible signal stored in an audible signal generation circuit is determined as data in which each audible signal is associated with each bit position. a first memory storing the data, a second memory storing a plurality of pieces of output data of the first memory read out each time a digital code word is received for each audible signal; a determination circuit that determines whether or not all corresponding digital code words from a plurality of pieces of stored data are present in a digital code sequence representing a specific audible signal; and an audible signal identified from the output of the determination circuit. and an encoding circuit that encodes and outputs the type of audible signal identification circuit.

The present invention will be described in detail below with reference to the drawings.

FIG. 2 shows an embodiment of the present invention, in which a first memory 40 determines whether a received digital code word applied to an input terminal 41 is included in a digital code sequence representing each audible signal. is stored in a memory area whose address is the digital code word. In the memory area, each bit is associated with each audible signal, and whether or not it is included in a digital code sequence is stored.

This will be explained in more detail with reference to FIG. 3 (A shows the audible signal waveform and code sequence, B shows the configuration of the first memory). In FIG. 3, the first audible signal consists of a digital code sequence of _n1 words, and the second audible signal consists of a digital code sequence of _n2 words. the first of the first audible signals
Since the word is 00000000, data "1" indicating that the address is included in the code sequence is stored in the bit position corresponding to the first audible signal in the memory area of the first memory address 00000000. . Similarly, since the second code word is 11010001,
A "1" is stored in the corresponding bit position of the memory area with this address. Similarly, n ₁
Store data for words. At this time, the data of code words that appear duplicated remains as "1", while data "0" is written in all corresponding bit positions in the memory area whose address is the code word that does not appear at all. Data is stored in exactly the same manner for the second audible signal.

If the first memory 40 storing data in this manner is read out using the received digital code word as an address, each bit position will contain the digital code word included in the code sequence of the corresponding audible signal. You can display whether there are any. Second
Memories 42, 43, and 44 are provided for each audible signal, take in data thus displayed each time a digital code word is received, and always hold m pieces of data. second memory 42,
43 and 44 are each composed of m-bit shift registers in the simplest form, but random access
It is also possible to realize an equivalent function using memory (RAM).

Judgment circuits 45, 46, and 47 each determine whether or not m pieces of data include all corresponding code words in the code sequence of the audible signal (third
In the example shown in the figure, it is determined whether all m pieces of data are 1. Based on these determination outputs, the encoding circuit 48 encodes the type of audible signal whose presence has been identified and outputs it to the output terminal 49.

As can be seen from the above description, the audible signal identification circuit according to the present invention has an extremely simple configuration. That is, there is no need for an expansion circuit for the first digital film or a decoding circuit for the analog filter. Further, the basic operation of the identification circuit is only reading from the memory and simple pattern detection on the read data, and does not perform any filter operation. Once the digital code sequence of each audible signal to be stored in the audible signal generation circuit is determined, the first
The data stored in the memory 40 of is automatically determined. Further, since no filter operation is performed, there is no need to design optimal filter orders and coefficients, and the circuit design is extremely easy. Also, even if the standards for audible signals change, the first
This can be easily handled by simply changing the data stored in the memory 40 of the computer. If the world standard 8-bit digital code is used, the memory capacity of the first memory is 256 words, 16Kbit,
In the recent situation where 32kbit memory has also appeared, it can be said that the memory requirement is a very small amount.

On the other hand, sufficient identification characteristics can also be obtained. The problem with discrimination characteristics is the ability to reliably identify the signal to be identified, the ability not to falsely identify speech signals, different audible signals. There are almost no problems with the former. That is, as is clear from the above description, when a signal to be identified comes in, the determination circuit 45 and the like can definitely output an identification signal as long as there is no code error. However, in that case, a circuit for determining an audible signal that should not be identified may also output an identification signal.

Some measures are required to avoid this problem and reduce the probability of misidentification of audio signals.

For example, when a dial tone (350Hz, 440Hz) is expressed as a code sequence of 91 words according to the μ law, one of the world standards (the frequency is exactly
351.6484Hz, 439.5604Hz), an example of the types of codewords appearing in the 91 words and the number of each codeword is shown in FIG. 4 (blanks represent 0). Because some codewords appear repeatedly, the number of codeword types that appear in 91 words is 63, which is about 1/2 of the 256 codeword types.
It is 4. Also, looking at the distribution of codes in FIG. 4, there are relatively few codes in which the signal level is low, that is, the upper 4 bits appear in hexadecimal representation of 0 to 3 or 8 to B. Since many codes of an audio signal are distributed in areas where the signal level is low, the probability of misidentification of the audio signal is considerably small. Even if it is assumed that the probability of the appearance of a code in a voice signal is uniform over all signal levels, the probability that that code will match the code present in the dial tone code sequence is approximately 1/4.
It is. If m = 8, when the code of the voice signal matches the code in the dial tone code sequence over 8 words, the voice signal will be misidentified, and the probability is P = (1/4) ⁸ = It can be extremely low as 1.5×10 ^-5 . If we consider that the codes are highly distributed in areas where the signal level is low, the probability of this happening is thought to be even lower.

In addition, by making use of the symmetry of the audible signal's waveform and introducing as much symmetry as possible into the codeword sequence, or by slightly changing the signal level within the standard, we can create an audible signal with as few types of codewords as possible. By configuring a code sequence of , it is possible to further reduce the probability of misidentification.

In addition, misidentification between audible signals can be adequately countered by selecting in advance code sequences such that the number of codes that appear together in each code sequence is as small as possible, and by selecting m necessary to prevent misidentification. It is possible.

As described above with respect to the embodiments, according to the present invention, an audible signal identification circuit with excellent characteristics can be obtained with a very simple configuration without using complex functions and circuits such as filters. The effect is extremely large.

The present invention is applicable when the audible signal is generated by periodically reading out a digital code sequence stored in a memory, but it is also applicable to push button (PB) signals, multifrequency (MF) signals, etc. is generated by a similar method, it can also be applied to PB signal receivers (PBR) and MF signal receivers (MFR), and the configuration is much simpler than conventional PBRs and MFRs. Therefore, the effect will be enormous.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram showing an audible signal identification circuit according to the prior art, Fig. 2 is an explanatory diagram showing an embodiment of the present invention, and Fig. 3 is an explanatory diagram showing an example of the present invention. FIG. 4 is a diagram showing an example of the distribution of code words appearing in the code sequence of an audible signal. In the figure, 11 is an expansion circuit, 12, 13, 1
4, 15, 20, 21, 22, 23, 26, 2
7, 28, 29 are filter circuits, 16, 17, 2
4, 25, 30, 31 are detection circuits, 18 is an identification logic circuit, 40 is a first memory, 42, 43, 44
is a second memory, 45, 46, 47 are determination circuits,
48 is an encoding circuit.

Claims (1)

[Claims] 1. In an audible signal identification circuit of a digital exchange having an audible signal generating circuit that generates an audible signal by periodically reading out a digital code series stored in a memory, a digital code word is used as an address, and the digital code word is an audible signal. a first memory storing data in which each audible signal corresponds to each bit position, indicating whether or not the audible signal exists in a digital code series representing each audible signal stored in the signal generating circuit; and a digital code word; a second memory that stores a plurality of pieces of output data of the first memory for each audible signal, which are read out each time an audible signal is received; and a corresponding digital code word from the plurality of pieces of data stored in the second memory. a determination circuit that determines whether all of the above are present in a digital code sequence representing a specific audible signal, and an encoding circuit that encodes and outputs the type of audible signal identified from the output of the determination circuit. An audible signal identification circuit comprising: