JPS5815990B2 - Signal sampling scheme for time division multiplexers - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a time division multiplexing device (hereinafter referred to as a TDM device) used for the purpose of allowing a plurality of communicators to exchange information using a single communication path, especially in a telex switching network. It concerns a signal sampling method for

Here, the TDM device described in this specification is a system that time-division multiplexes four telegrams with a communication speed of 50 bits/second from multiple communicators to generate a voice frequency of 0.3 KH,z to 3.4 KHz. It means a device that performs communication by converting the communication speed to a high speed, such as 1200 bits/second, 2400 bits/second, 4800 bits/second, or 9600 bits/second, through a band signal transmission path.

Conventionally, in the transmission of this type of telegram No. 4, the band was divided into smaller frequency bands and allocated to each communicating party, so that each communicating party communicated using only the assigned frequency band. Frequency division multiplexing equipment has been widely used.

However, in recent years, there has been a frequent demand for the use of TDM devices that have high multiplexing efficiency and can economically use communication networks.

However, in the case of frequency division multiplexing, each communicating party can communicate independently, whereas in the case of time division multiplexing, the use of the communication network is allocated to each communicating party by time division. It becomes necessary for the temporal relationship found in the telegraph story to be established.

However, the signal called the start-stop signal from each communication party used in current TDM equipment is a start bit (corresponding to the "1" level of the A polarity binary logic level) as shown in Figure 1. ) starts with a stop bit (Z polarity; 2
Within a character ending with a logic value (corresponding to the "0" level), the relationship between the first conversion point of the start bit and other conversion points within that character is always an integral multiple of the reciprocal of the communication speed. However, it is a signal in which no temporal relationship is established between each character.

The relationship between conversion points within one character (meaning conversion from A polarity to Z polarity or from Z polarity to A polarity) is a nominal one, and in reality, the positional relationship between each conversion point is the nominal position. Distortion is occurring due to the deviation.

Therefore, in order to transmit such start-stop signals from each communication party using a fixed time period, it is necessary to synchronize them according to a fixed relationship.

As shown in Figure 1, this type of synchronization method first detects the start bit of the input start-stop signal and counts the 1/2 bit length of time from the first conversion point (falling edge) of the start bit. The first sampling pulse is generated at a time point, and the second to seventh sampling pulses are sequentially generated at each point in time when the time of 1 bit length of the nominal data rate of the start-stop signal is counted, resulting in seven samplings for one character. A method has been considered in which each sample value obtained by sampling pulses is extracted in accordance with a synchronized signal.

Hereinafter, the signal synchronized in this manner will be referred to as an aggregate signal, and the pulse for extracting sample values as a synchronized signal will be referred to as an aggregate sampling pulse.

Sampling the input start-stop signal in the time series of sampling pulses as described above provides the maximum margin for distortion of the start-stop signal (up to a maximum of about 46 coefficients), and this can be considered in terms of temporal timing. For example, if the start-stop signal is input without distortion at the nominal data rate, it means that the sampling pulse is at the temporal center position of each bit of the start-stop input signal. There is.

It should be noted here that the above sampling method cannot be used as a sampling pulse for a call setup signal as described below in a telex switching network.

In particular, regarding the telex exchange network, as mentioned above.

In addition to the start-stop signal, the following four signals are sent and received to establish a communication path between the exchange and the opposing TDM device.
In this case, various control signals (hereinafter collectively referred to as call setup signals) are required.

That is, the call setup signal used for this purpose is C@C
-1, T, T, (International Tel
egr-aph and Te1ephone Co
(abbreviation for n5ultative co-mm i ttee) recommendation document, it is roughly divided into four types: type A signal, type B signal, type C signal, and type B signal, and each of these signals includes an idle line as described later. This includes signals, calling signals, connection confirmation signals, selection signals, etc.

The outline of the control procedure using these signals for type A signals and type B signals is as follows.

In other words, when a communicating party uses its TDM device to exchange information with another communicating party's TDM device via a switching network, initially both TDM devices are not connected to the communication path and send out idle line signals to the exchange. are doing.

In this state, the calling side that wishes to communicate first sends out a paging signal.

Then, the exchange receives this and continuously sends a paging confirmation signal and a selection start signal to the calling party in place of the idle line signal.

The calling party then sends out a selection signal in response.

Upon receiving this selection signal, the exchange searches for a TDIVI device on the other party, that is, the called side, secures a communication path, and sends a connection completion signal.

This starts information exchange between both TDM devices.

When the information exchange is completed, the calling party sends a recovery signal.

In response to this, the called side sends out a recovery confirmation signal, after which both TDM devices become idle again.

If the paging confirmation signal is not received within certain conditions, a retest signal is sent from the calling side and the called side is exchanging information with a TDM device other than the calling party. etc., a service signal called a busy signal is sent from the exchange to the calling party.

These call setup signals have various time formats depending on the type, and among them, the dial pulse, which is the selection signal in the Quip B signal, is mechanically generated by a method equivalent to dialing in a normal telephone. The pulse train (C, C, 1
, T, T, shown in Section 0.2 of the Recommendation) and is different from the normal start-stop signal as described above.

On the other hand, the control procedure for call setting signals in type C signals and type D signals is almost the same as in the above-mentioned type A and type B signals, but in the case of type C signals and type D signals, the calling side sends out the paging signal. This procedure is different from those for type A signals and type B signals in that the selection signal is sent out continuously without receiving a paging confirmation signal from the post-exchange.

Type A signals, type B signals, and type C signals are divided into the calling side and the called side, and time conditions are attached to the calling signals, calling confirmation signals, etc., and according to the present invention, the third part, which will be described in more detail later, 4 and 5, respectively.

When a TDM device is used in a telex switching network, it is important to reliably sample and synchronize the start-stop signal and call setup signal as described above, and to be able to send them out onto a communication channel at a high communication speed.

Then, the TDM device on the receiving side needs to retransmit this signal to the other device as a signal within internationally defined standards.

Does that mean 'L? This is because the signal transmission method in the DM device is such that the input signal is sampled and the sample values obtained are sent out as a pulse train at constant time intervals.

Therefore, in the process of processing such signals, a particular problem is how accurately and reliably the sampling is performed.

In particular, it may not be possible to obtain an accurate sampled value for a certain signal in the call setup signal by using only the sampling method as shown in the explanation of the start-stop signal above.

Therefore, in conventional TI)M devices, as a means to solve this problem, (1) when a conversion point from Z polarity to A polarity (start bit conversion point) is detected, a sampling pulse for one character (usually 7 (2) A method that generates a sampling pulse of 4 bits when a transition point from A polarity to Z polarity is detected;
After detecting the point of conversion to polarity, sample one character.

When a pulse is generated and a signal is constantly being sent with A polarity, a 4-bit sampling pulse is generated at the first transition point from A polarity to Z polarity, and the signal is transmitted between the receiving system and the transmitting system. If A polarity constantly appears on either side, a typical method is to detect the transition point from continuous transmission of Z polarity to A polarity and then generate sampling pulses for 4 bits each. It is considered as something.

However, in the former method 1, the range of distortion allowed for dial pulses is small, and even when dial pulses are regenerated, it is not detected that they are dial pulses, so dial pulses are transmitted accurately. There are some drawbacks that cannot be avoided.

On the other hand, in the latter method 2, when transmitting Type C signals and Quip D signals, the transmission of A polarity is detected in the transmitting and receiving system to detect that a dial pulse is being transmitted. There is a drawback that the rear start-stop signal may be erroneously sampled as a dial pulse (see FIG. 5).

That is, if four sampling pulses are used when transmitting a dial pulse, and the dial pulse speed is 10 pulses/second - 10%, the width of the pulse (A polarity) is 70 m5.
In some cases, the sample value exceeds the value, and accurate sample values cannot be obtained.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a signal sampling method for a time division multiplexing device that can overcome the above-mentioned drawbacks of the prior art.

That is, if we focus on the call confirmation signal and selection confirmation signal in the type B signal, if these signals are detected in either the transmission or reception system of the TDM device, a dial pulse will be transmitted to the opposite system. become.

Therefore, the present invention has been made with the focus on the fact that all of the above-mentioned problems of the conventional art can be solved by effectively utilizing this relationship to control the generation state of the sampling pulse.

Hereinafter, an embodiment of a signal sampling method for a time division multiplexing apparatus according to the present invention will be described in detail with reference to FIGS. 2 to 5.

First, prior to a detailed description of the present invention, signal sampling conditions and regeneration conditions in the present invention will be listed.

a) Conversion point of the start bit of the start-stop signal (from A polarity to Z
After detecting a signal conversion point (conversion point from Z polarity to A polarity and conversion point from A polarity to Z polarity), a certain monitoring time is set to ensure that this conversion point is valid. and whether it is valid as a start bit.

If valid, the channel input of the telex line is sampled by selectively generating two sequences of sampling pulses as described below.

b) One of the two types of sampling pulse sequences mentioned in section a above is a sequence of seven sampling pulses that can be adapted to the start-stop signal (also called start-stop signal) commonly used in telex lines. one sampling pulse, the other one matching the control signal used to secure the line.

C) The selection of the two types of sampling pulse sequences as described above is
This is done by four types of control signals generated inside the TDIVI device.

These four types of control signals are 5SPS and R8PS.

The sequence of seven sampling pulses, tentatively named SDF and RI)F, is selected when RDF is at logic "0" level and 5SPS is at logic "0" level in the transmitting system, and SDF is selected at logic "0" level in the receiving system. , J level and when R8PS is at a logic "0" level.

On the other hand, one sampling pulse sequence is generated when either RDF or 5SPS is at logic "1" level in the transmitting system, and when either SDF or R8PS is at logic "1" level in the receiving system. Each is selected in certain cases.

However, the sequence of sampling pulses in the receiving system does not actually involve sampling, since the input is synchronized with the speed on the line, but is equivalent to latching.

Therefore, in the case of a receiving system, this will be referred to as a latch pulse sequence hereinafter.

a) The SSPS signal condition is that if the signal on the terminal side is detected to carry Z polarity of 135 m8 or more counting from the point of conversion from A polarity to Z polarity, it becomes a logic "0"level; A of 135m8 or more counting from the point of conversion to polarity
A logic "1" level occurs when polarity transport is detected.

In other signal states, the previous state is maintained.

e) The R8PS signal condition is that the signal of the corresponding channel on the line side has 135 points counted from the point of conversion from A polarity to Z polarity.
If transport of Z polarity of m8 or more is detected, the logic becomes "0" level, and conversely, if transport of :A polarity of 135 m8 or more counting from the point of conversion from Z polarity to A polarity is detected, the logic becomes "0".
1” level.

In other signal states, the previous state is maintained.

f) The SDF signal conditions are such that a short pulse of Z polarity (17 to 35 m
5 width, which corresponds to the call confirmation signal and selection start signal in the Quip B signal) is detected, the logic becomes "1".
level, and then inverts to logic "o" level with a 5SPS signal when detecting Z-polarity transport of 1.35m5 or more.

In other signal states, the previous state is maintained.

g) The RDF signal condition is that A polarity transport is detected in the signal of the corresponding channel on the line side.

(that is, when R8PS is at logic “1” level), Z
Short pulses of polarity (20 mS or 40 m5 width,
This width was 17 to 35 m5, but it became this width because it was synchronized with the opposite TDM device), it becomes a logic "1" level, and then 135 m
When a Z polarity of 8 or more is detected, it is inverted to a logic "0" level together with the R8PS signal.

In other signal states, the previous state is maintained.

h) One sampling pulse sequence is generated at the conversion point regardless of the rising or falling edge of the channel input signal from the telex line.

At this time, if the input signal is an impulsive pulse (pulse width not exceeding 0.5 bit width), no sampling pulse is generated.

i) The seven sampling pulse sequences are generated at the falling edge of the channel input signal, that is, when converting from Z polarity to A polarity, and are not generated at the rising transition point.

J) The intervals between the two types of sampling pulse sequences generated in the transmission system are as follows.

That is, if the channel input signal is a start-stop signal, the first sampling pulse is generated at a time position approximately 1/2 bit after the first conversion point of the start bit, and the other six sampling pulses are then generated at each time point. This occurs continuously at intervals of one bit length.

In addition, if the channel input signal is deemed to be other than a start-stop signal, one sampling pulse is generated 10m5 after the conversion point for each conversion point from A polarity to Z polarity and from Z polarity to A polarity. .

Next, the block diagram of FIG. 2 showing an embodiment of the method of the present invention will be explained.

b) The channel input of the block telex line on the transmission system side (CH manual input in the figure) is first input to the sampling block 11.

The sampling block 11 performs high-level sampling on this signal using a clock CKI in the figure, and inputs the sampled value obtained thereby to the data processing block 12.

The data processing block 12 receives this sample value and uses the signal 5SPS from the transport monitor block 13 and the signal RDF from the dial monitor block 25 in the receiving system block to be described later to select the above-mentioned sampling pulse sequence. and perform the occurrence.

In addition, detection of the start bit in the normal start-stop signal,
Disposal of start bits not exceeding 0.5 bits, impulse nozzle generated during continuous feeding of A polarity and Z polarity (
(not exceeding 0.5 bits wide), RAM block 1
It generates control signals for writing and reading data in RAM block 14, and selects writing and reading addresses for the RAM block 14.

The sequence of one or seven sampling pulses generated by the data processing block 12 as described above is then:
It is sent to the sampling block 11 as a sampling clock CK2, and the sample value sampled by this sampling pulse is input from the signal output terminal 801 of the sampling block 11 to the input terminal DI of the RAM block 14.

Reference numeral 15 in the figure is a dial monitor block that acts as a detection circuit for generating the signal SDF.
A call confirmation signal and a selection start signal of type B signals used in a telex line are detected under the conditions shown in the above section f.

The transport monitor block 13 is configured to carry out continuous transport of Z polarity and
This block is for detecting continuous transmission of polarity, and operates to receive various signals from the data processing block 12 and generate 5SPS under the conditions described in section d above.

Furthermore, in the RAM block 14, the sampling block 11
Under the control of the address signal from the data processing block 12 and the read/write control signal R/W, the corresponding R
Acts to write into AM block 14.

The output Do of this RAM block 14 is outputted from the RAM block 14 under the control of the address signal and R/W signal from the data processing block 12, and then becomes a signal on the transmission path by the action of the launch 16 at the next stage. It is read into the application gate under the control of the clock CK3 synchronized with the application gate and becomes the application gate upper number.

Here, the RAM block 14 is a FIFO (First In First Array).

It consists of a memory circuit in the form of 1.5 and 1.3 and functions to reproduce asynchronous signals from the telex line as synchronous signals on the application gate.

口) Receiving side block Roughly speaking, the receiving system is the opposite of the transmitting system described above.

However, since the input signal is obtained as a synchronizing signal, transmission No. 4 is regenerated by latching it with the latch 27 and providing it as a channel output.

However, this simple operation does not guarantee the stop bit length of dial pulses or normal start/stop signals.

Therefore, an output monitor block 28 is added to ensure the standard of the dial pulse and the stop bit length of the start/stop signal.

Incidentally, on the receiving system side, a sampling block like that on the transmitting system side is naturally unnecessary, and the above-mentioned data processing block 12, transportation monitor block 13, and RAM on the receiving system side
Parts corresponding to the block 14, dial monitor block 15 and latch 16 have reference numerals 22.23.2, respectively.
4.25 and 26 are attached and detailed explanation thereof is omitted.

Since the output monitor block 28 has a synchronous system in which the stop bit of the start-stop signal and the pulse width of the dial pulse are synchronously sent from the application gate, the stop bit is 1 bit long. The length is 2 bits or more, and the width of the dial pulse may be 2 bits long, 3 bits long, or 4 bits long, so
To ensure this, the output launch pulse series (CK4) and RA are transmitted and received with the data processing block 22.
The timing of the M function cook 24 dress signal and R/W signal is temporally manipulated so that the width of the dial pulse and the stop bit length of the start/stop signal fall within the standards.

However, as can be easily understood from the above, the receiving system dial monitor block 25 is for generating the signal RDF, and the transportation monitor block 23 is for generating the signal RDF.
It is for generating 8PS, and it goes without saying that it is different from the dial monitor block 15 and continuous transmission monitor block 13 on the transmission system side in this respect.

The effects obtained by using the above-described sampling method will be explained in detail using FIGS. 3 to 5 as examples.

First, regarding the type A signal shown in FIG.
sps inverts from a logic "1" level to a "0" level.

Here, the signal from the called side includes a short Z-polarity pulse (1
7 to 35 ms) is not sent, the signal RDF
maintains a logic "0" level state.

Therefore, the next time the calling side sends out a selection signal (start-stop signal), the signal 5SPS will be set to logic ``0 level'' and the signal RD
Since F is also at the logic "0" level, a series of seven sampling pulses for the start-stop signal is generated.

Regarding the type B signal shown in Fig. 4, the signal 5SPS is no longer valid after a monitoring time of 135m5 from the selection of the calling signal.
l [Reverses from 1,4 level to "0" level,
Since a short width pulse of Z polarity, that is, a calling confirmation signal and a selection start signal are sent to the signal from the called side, the signal RDF is inverted to the logic "1" level.

Therefore, a selection signal (dial pulse) is sent from the calling party next.
When the signal 5sps is sent out, the signal 5sps is at the logic "0" level and the signal RDF is at the logic "1" level, so one sampling pulse is generated.

In addition, even when Guyarnolus occurs again, the control signal of the transmission system on the called side is logic "1" and the processing system is operating as if the dial pulse is being input.
It is possible to keep the dial pulse train within specifications.

Here, the signal SDF of the transmission system of the called side and the signal RDF of the reception system of the calling side are because the transmission of the dial pulse is completed and the connection completion signal (carrying Z polarity for 2 seconds or more) is sent from the called side. The logic is reversed from the ``moon'' level to the ``0'' level and does not affect the subsequent transmission and reception of signals.

In the case of the type C signal shown in FIG. 5, the call signal includes a start-stop signal.

That is, after sending out the Z polarity of 150 m8 to 300 m5, a start-stop signal with a predetermined sign is sent out, and this is followed by a selection signal (start-stop signal).

In this case, 135 m5 from the start of sending the calling signal.
After the monitoring time of , the signal 5SPS inverts from the logic "1" level to the "0" level.

Since the called side does not send out a short pulse of Z polarity and is in an idle line state (carrying A polarity), signal RDF is at a logic "0" level.

Therefore, when the start-stop signal in the calling signal or the selection signal (start-stop signal) is sent out, the signal 5SPS is set to logic rOJ.
The level of the signal RDF is also at the logic [0, 4 level, and seven sampling pulses for the start-stop signal are generated.

Note that the type D signal has almost the same format as the type C signal described above.

In addition to these signals, there are retest signals, start-stop signals for exchanging data, etc., and these can be adapted to the present method.

It will therefore be clear from the foregoing detailed description that the sampling scheme of the present invention can be used in telex switching networks using all types of call set-up signals without the need for additional intervention, especially dial pulses. It has the advantage that it can be regenerated as a pulse almost within the specifications.

Therefore, by using a circuit constructed by applying the sampling method of the present invention, there is no need for the TDM device user to use different TDM devices depending on the type of call setting signal used by the Tl)M device user. Advantageously, the permissible distortion for the setting signal, particularly the dial pulse, is wider than that of conventional devices.

Note that the time charts for the type A signal and type C signal shown in FIGS.
Since F is always output at a constant level, no control is required.

Therefore, it should be noted that only the transmission system time chart of the calling side device and the transmission system time chart of the called side device are shown respectively.

In addition, this time chart only takes into account the delay that occurs during sampling and the delay that occurs when synchronizing the signal on the application gate.Delay elements may occur for various other reasons, but in actual use, this will not occur. It is judged that this does not pose a problem.

Furthermore, only one of the signals marked * in FIGS. 4 and 5 is generated when an answer is actually given.

[Brief explanation of the drawing]

Fig. 1 is a diagram shown to explain the basic sampling method of a start-stop signal that has been used conventionally, Fig. 2 is a block diagram of an example circuit using the sampling method according to the present invention, and Fig. 3 , FIGS. 4 and 5 show the second control procedure for typical control procedures for type A, type B, and type C call setup signals, respectively.
FIG. 3 is a diagram displaying the response shown by the circuit example shown in the figure as a time chart.

Claims (1)

[Claims] 1 In time division multiplexing equipment used in telex switching networks, type A, type B,
A circuit for detecting the transport of A polarity of an input signal for a certain time length or more in each of the receiving system and the transmitting system for the purpose of sampling in accordance with four types of control signals consisting of type C and type D; It is equipped with a circuit that detects Z polarity transport for a certain period of time or more, and a circuit that detects a call confirmation signal and a selection start signal in the type B signal, and uses the circuit outputs that are the results of these detections to generate a start-stop signal. The input signal of the telegraph channel can be sampled by switching between the time series of seven sampling pulses for sampling the input signal and one sampling pulse for sampling the dialing signal, ringing signal, etc. A signal sampling method for a time division multiplexer, characterized in that: