JPS60109940A - Optical multiplex transmission method and its system - Google Patents

Abstract

PURPOSE:To attain accurate transmission even any environment by attaining four kinds of checks such as pulse width check at the reception side of the optical multiplex transmission system. CONSTITUTION:An optical signal from an optical cable 19 is converted into an electric signal at a photoelectric converting section 9, given to a code identification section 10 and the pulse width is checked by a pulse width check section 11 provided to the identification section 10. In identifying a code, the code is divided minutely, 8 blocks at each 4 bits, 3 bits out of 4 bits in one block are used and then H, L are discriminated, so-called a majority decision system is adopted to attain stable code identification even under a tight noise environment. At the post-stages, three check functions are provided as pulse number check, a parity check section 13 and a summing check section 17a thereby attaining respective checks.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention allows signals of up to 256 points that change relatively slowly, such as contact signals and measurement control information, to be processed as they are, or as an A/
An optical multiplexing-transmission method that inputs data to a transmitter via a D interface, transmits it to a receiver via a time-division multiplexed optical fiber cable, and enables accurate transmission under any environment. It's about that system.

Conventionally, systems that use metal wires for the transmission of contact signals or measurement control information, etc. are defenseless against noise that invades the transmission line, and it is difficult to identify the type and size of foreign noise. However, it is impossible to take sufficient check measures against any noise environment, and there is a major drawback that signal transmission may be completely impossible depending on the noise environment.

Therefore, there are systems that use optical fibers for power transmission lines, focusing on the excellent electromagnetic induction resistance of optical fibers, but in such systems, it is unlikely that foreign noise will enter from the transmission line of the optical fiber device itself. This is a major problem in that optical fiber's excellent resistance to electromagnetic induction cannot be fully utilized because there is no thorough check for noise transmitted through the power supply and noise that enters from the electrical signal system at end stations such as transmitters and receivers. It does have its shortcomings.

However, in the present invention, since an optical fiber cable is used for the transmission path, the locations where noise enters are limited as described above, and it is also easy to estimate the amount of noise that enters. A total of four types of checks are adopted: a pulse number check, a parity check, and a double match check using an automatic transportation system. This makes it possible to fully utilize the excellent electromagnetic induction resistance of optical fibers, ensuring accurate results in any environment. This made transmission possible.

Reference will now be made in detail to the accompanying drawings that illustrate one example of the implementation of the invention.

l is a large number of input stage latches, each at the input end of the 8-tree,
A contact signal from a contact or an analog signal from a measuring device is connected through an A/D interface through a photocoupler 2 that emits a no-voltage contact signal.There are a maximum of 32 contact signals and a maximum of 256 contact signals and a measuring device. It inputs the analog signal of Here, the photocoupler 2 has the effect of protecting the circuit from surges and large induced currents. Reference numeral 3 denotes a control section which is connected to the clock pulse input pin 3a of each input stage latch l, and provides a time difference between the operations of all input stage latches l.

4 is a timer, which is connected to the control unit 3 and connected in parallel with the output pin of the tree that supplies the clock pulse that is the reference for the time division, and the parallel data latched at the same time by each input stage latch 1. signal, control section 2
This circuit converts data into a serial data signal by sequentially transmitting data from each input stage latch 1 with a time difference.

Reference numeral 6 denotes a long/short code conversion and start/stop/parity code addition unit connected to the output of the parallel/serial conversion unit 5, which is connected to the control unit 3 and is connected to the output of the serial data signal as shown in FIG. Start signal ST as shown
, later, a parity bit P and a stop signal SP are added to make one frame, and at that time, only the stop signal SP is changed between the first and second frames, and the first time is SPI and the second time is SF3. It is configured so that At this time, the timing at which each input stage latch l latches the input is when each signal of the two frames has been sent out and the signal from the control unit 3 is received.
The period (sample jig interval) is approximately 2 as shown in Figure 3.
1 m5ec (10,36X 2 = 20.72m5
ec).

In addition, at the same time as the frame is constructed, the length code converter 6a converts the serial data signal to ensure code identification.
(Ll has a pulse rising width of 020μ at the beginning in the code (4θμ crown), and 111 has a pulse width of 10fisec in the code (4θμ crown), which is a known long and short pulse code (binary PWH = pulse width modulation). ) method. Reference numeral 7 denotes a photoelectric converter, which is composed of an optical digital link module, and converts the electric signal from the code converter and code adder 6 into an optical signal suitable for the optical fiber. The above constitutes the transmitter 8. 9 is a similar photoelectric conversion section which receives the optical signal from the photoelectric conversion section 7 and converts it into a long/short code string of electrical signals. .10 is a code identification unit connected to the output end of the photoelectric conversion unit 9, and is used to control the data signal D1 start signal ST and stop signal 8SP of the two frames converted into the long and short code strings. No. and Hibari Tabit P
This is to identify the check signal. Further, the code identification unit 10 is provided with a pulse width check unit 11 composed of four 8-bit shift registers, etc., and one code (4θμ5
Divide ec) into 32 parts (32 bits) and use 8 bits each to determine H (High) or L (j, o) by majority vote.
w).

That is, the determination is made based on whether two or more of the three bits are H or L, and the remaining one bit is not used in order to provide tolerance for timing deviations, etc. "01
11 “Ill l” STl+ I spx,, l
``The sign of SP2+1 is determined by H,
The combination of L is read, and if the combination does not correspond to any code, it is skipped. 12
2 is a control section which inputs the data signal and control signal of the code identification section 1°. 13 is a pulse number check and parity check section, and a code identification section 1o
This is to input the check signal. 14,14,1
Reference numeral 4.14 denotes a serial/parallel conversion section, which is connected to the output of the control section 12 and converts the data signal of the code identification section 1o into temporally parallel signals. Reference numeral 15 denotes a large number of first data latch sections, each of which is connected to the output power of the serial/parallel converter 14, and each serial/parallel converter 14 is provided with 8 of them, for a maximum of 32 latch sections.

Reference numeral 16 denotes the same number of output stage latches as the first data latch section 15, and a maximum of 25
It outputs an analog signal from a measuring instrument via a six-point contact signal or a D/A interface. 17
is the butt check section and latch control section,
It is connected to the pulse number check and parity check section 13 and interposed between the input and output terminals of each -th data latch section 15, and is connected to the latch control section 17b and each output stage latch 16. This constitutes the receiver 18. 19 is an optical fiber cable, transmitter 8
The photoelectric conversion section 7 of the receiver 18 and the photoelectric conversion section 9 of the receiver 18 are connected, and an optical multiplex transmission system 20 is configured. Here, the optical signal from the transmitter 8 received by the receiver 18 is converted into a long/short code string by an electric signal in the photoelectric converter 9, and the data signal is converted into serial/parallel data via the code identification unit 1O. If the stop signal sp of the parallel signal is the first signal SPI, the data signal is latched by the first data latch unit 15, and then the second signal is j′D for data transmission
is similarly converted into a parallel signal, but the second data signal is compared with the data signal latched by the first declutch section 15 in the matching check section 17a, and only the data is compared to see if it is correct. , not latched. - When the matching check unit 17a confirms that the first and second data signals are exactly the same, the data signal latched by the first data latch unit 15 is transferred to the output stage by a signal from the latch control unit 17b. It becomes an output signal only after it is latched by the latch 16. However, there are differences in the data of the two times in the matching check section 1.
7a, or the pulse width check unit 11
When the code is skipped, the pulse number check and parity check section 13 outputs a Sheller signal! If so, the data is not latched in the output stage latch 16 and is discarded. In this case, the output stage latch 16 is configured to bold the previous data until the next new data arrives, and at the same time turn on an error lamp (not shown) attached to the receiving JIA 18.

The data signals latched in each output stage launch 16 are input to the last stage seven-stage Oto coupler 2, and the output functions as an open collector 21 similar to a contact point. Fifth
The figure shows an example of a basic configuration diagram in which the optical multiplex transmission system 20 is used for centralized management of buildings, monitoring FIF control of factory power equipment, measurement control of factories, etc. A transmitter 8 and a receiver 18 are installed in both sides of the monitoring room 23, and the transmitter 8 and receiver 18 in a building, factory, etc. are connected to the receiver 1 in the central control room or the central monitoring room 23, respectively.
8 and the transmitter 8 are connected by optical fiber cables 19 and 19. 24 is a contact signal from a contact, 25 is an analog signal from a measuring instrument, and 1. Contact signal 24 is the same \analog signal 25 is analog/digital (A/D)
The signal is input to a transmitter 8 in a building, factory, etc. 22 via an interface 26. 27 is a display device installed in the central control room or central monitoring room 23, and the receiver 18 of the room 23
directly or digital/analog CD at the output end of the
/A) It is connected via the interface 28. Reference numeral 29 denotes a control device, which is installed in the room 23 and connected to the input side of the transmitter 80 in the same room 23. 30 is a controlled device such as a switch, which is equipped with 22Vcg in buildings, factories, etc.
It is connected to receivers 18 in 22 buildings, factories, etc.

The present invention has the above configuration, and the pulse width check is performed by the pulse width check section 11 attached to the code identification section 10 of the receiver 18, and in order to ensure code identification as described above, a long/short pulse code method is used. We are hiring. Then, to identify the code, one code (40μ5ec) is subdivided (3
(divided into 2 - 32 bits), divided into 8 blocks of 4 bits each, and used 3 bits of the 4 bits in 1 block to set H and L.
This system employs a so-called majority voting method to determine the code, and enables stable code identification even in a highly noisy environment. The timing to judge the code is when the first pit of the 32-bit shift register becomes H, and even if a normal code is read, the pulse number check and the pulse number check of the parity check section 13 are always performed. This will not result in sending out incorrect information.

Here, we will discuss the degree of improvement in the error rate using majority logic.

The error rate pb of one block is expressed as follows, assuming that the error rate iPa of one bit (1.25μ5ec) of the code identification shift register is expressed as follows.

pb = 8a□x (Pa)2x (1-pa) +
If the error rate of the pa8- code is pc, then the - code is composed of 8 blocks as described above, so Pc = l - (l - Pb) = 8Pb - 28Pb + 56Pb - 70Pb + 5ap
b -28Pb+8pb -pb, and by substituting the error rate of 1O-6, which is generally a standard in electronic circuits, into pa, Pb = 3X1
0. Pc = 2.4XlO'', a significant improvement can be obtained.

Next, the number of pulses is checked by the pulse number check and parity check section 13 connected to the code identification section 10 of the receiver 1a18, and the data signal is checked between the arrival of the start signal ST and the arrival of the stop signal sp. By counting the number of pulses, errors caused by signal skipping are checked.

Here, the number of data in one frame is 64°128 depending on the specifications.
．． Either 192 or 256 is taken, but if the waveform is distorted at the other end due to noise, there is a risk that one or more signals may be skipped. In that case, in the pulse number check and parity check exa, the number of data in the - frame is decreased and counted, and at this time it is determined that there was an error in the sign determination (pulse width check unit 11), and the data in that frame is not used. , hold the previous data as is. Therefore, this check function has a close relationship with the pulse width check section 11, and errors that pass through these two checks can be detected by "setting o++ to Ill, . . ." to "l2. The only error would be misreading 11 Q n.

Here, “011” is “”++’! Or’l1. is 11 Q
6. If pd is the probability of misreading, then Pd = (1-Pb)'x pb, where pb is the error rate of 1 block, and as mentioned above < P
Since b = 3xlO, Pd = 9 x 10''.

Furthermore, the parity check is performed by the pulse failure check and parity check unit 13 as described above.The parity check method adopted in the present invention is called vertical parity in JIS, and the
This is a method of checking by reading Ce1A. In this case - a pacity bit P is added to the end of the data signal of a frame, but this signal is added to the II l u of one frame.
The change in "Lt" of the O number is always observed as an even number. Now, if an error occurs in the signal due to the influence of noise, etc., at the erroneous location, 11 Q,, is ゛"++ straddle 12. is 011
and - read, 12. If the total number of changes is 1 due to an error in the frame, then 1. .

(2) Therefore, since the parity signal is also an odd number, it is detected that there is an error in the frame.

However, if the cause of the error is two or more even numbers, it cannot be detected, so this parity check is a method for detecting a single error.

Then, pe =2. a2x(pd)2x (1-pd)26'
As shown above, Pd = 9XlO, so P
e = 2.66 x 10'', so the probability of making three or more mistakes is extremely small, so we ignore it.

Finally, the double butt and alignment check by the automatic transportation method is performed by the alignment check unit and the alignment check unit 17a of the latch control unit 17, and this check is the final check of the present invention, and is the same for each frame. A signal is sent twice each time, and errors are detected by comparing the two signals. Even if there were an error in which the previous two checks were missed in a particularly noisy environment, the probability that such an error would occur twice in a row is extremely small. Verification is performed by comparing the signals of 256 points in one frame all at once, so an error that can pass this check is only possible if the exact same error occurs twice in a row at the same location. do not have.

Let Pf be the probability of the same error occurring twice in a row when there are 256 points of data signals in one frame. At that time, there are two errors including the parity check, so Pf is It is expressed as follows.

=256X255÷2XPd X(1-Pd) 'However, Pd is "02." or 12. The probability of misreading ``02'' as ``02'' is pf = 2.14xlO'' since <Pd=9X10'' as described above, and the probability of three or more errors is extremely small, so it can be ignored.

When the above four types of checking methods are used together, the error rate is set at 1O-6, which is a standard value for each electronic circuit, but as you can see, the error rate in the steady state is If the four types of check methods are used together, it will almost always be OK. However, in actual dynamic transmission systems, the code error rate can be extremely high instantaneously due to various environments. In optical transmission systems, which are often subject to various conditions, the dynamic characteristics under such conditions are important, and the combination of the four-way check method in the present invention has a great effect even in such worst environments. Demonstrate.

In addition, as mentioned above, in an optical transmission system, it is unlikely that foreign noise will enter from the transmission path, so the worst environment is considered to be an environment where noise is transmitted through the power supply and circulated.
Particularly large noise sources include arc discharge and the like.

In the case of arc discharge, several pulse noises occur during one cycle of the AC power supply.

Now, this noise occurs during the power supply (60Hz'z) cycle, but what happens? The error rate pa is Pa = 1.25xlOx60. x100=7.5
It becomes X10-8. At this time, the error rate Pb for one block is Pb = 3(Pa) -2(pa) = 1.68X10
' becomes. Next °02. “Il+ or IIl,
The probability pd of misreading , as “°02.” is pa = (1-pb) x pb = 2.82 x
10'', the probability of making the same error twice in a row, that is, the probability of not getting caught in any of the four types of checks, is pf = g56 x 255 ÷ 2 X Pd
)""=2.1xlO-26. This is the probability that any of the 256 signals will have an error after sampling - times, so the sampling interval should be set to 20m5ll! If c, then the average time t at which - times of errors occur is t, = 2xlO÷Pf -3,0x1016 (years), and it is considered that there are virtually no errors.

[Brief explanation of the drawing]

The attached drawings show an example of the implementation of the present invention. Fig. 1 is a block wiring diagram showing an outline of a transmitter, Fig. 2 is a block wiring diagram showing an outline of a receiver, and Fig. 3 is a block wiring diagram showing an outline of a receiver.・This shows a frame composed of a stop parity code addition section, where A is the first one, B is the second one, Figure 4 is a photocoupler, and Figure 5 is the optical multiplex transmission system of the present invention. This shows an example of a basic configuration diagram using . l...Input stage latch, 2...Photocoupler, 3...
- Control unit, 3a... Tarock pulse input pin, 4... Timer, 5... Parallel/serial converter, 6... Long/short code conversion and start/stop/parity code addition unit, 6. a... Long/short code conversion unit, 7.
... Photoelectric conversion section, 8... Transmitter, 9... Photoelectric conversion section, 10... Code identification section, 11... Pulse width check section, 12... Control section, 13... Pulse Number check and parity check section, 14... Serial
Parallel conversion unit, 15th...-th data latch unit, 1
6... Output stage latch, 17... Match check section and latch control section, 17a... Match check section, 17b... Latch control section, 18.
...Reception i, 19...Optical fiber cable, 20...
・Optical multiplex transmission system, 21...Open collector, 2
2... Pill, factory, etc., 23... Central control room or central monitoring room, 24... Contact signal, 25... Analog signal, 26... Analog/digital interface, 27... Display device, 28...Tystal analog interface, 29...Control device, 30...
・Controlled device, D...data signal, ST...start signal, SP, SPI, SP2...struff signal,
P... Paritypid. Applicant: Kinki Electric Works Co., Ltd. (and 2 others)

Claims (1)

[Claims] A/D-converted digital signal is transmitted as an optical signal from a transmitter to a receiver via a time-division multiplexing type optical fiber, and the optical signal is converted into an electrical signal to check the pulse width. An optical multiplex transmission method characterized by carrying out a total of 4 meters of checking, including a pulse number check, a parity check, and a two-time matching check using an automatic transport system. 2. Connect up to 256 points of contact directly or with an A/D interface for measurement control information, etc. to the input of the photocoupler, connect the input stage latch to the output of each photocoupler, and connect the clock of each input stage latch. Connect the control section that inputs Talokhals from the timer to the pulse input pin, connect the parallel/serial converter to the output of each input stage latch, and connect the control section to the output of the parallel/serial converter. A transmitter that converts a time-division multiplexed electrical signal into an optical signal and transmits the signal by connecting a code conversion and start/stop/parity code addition section and a photoelectric conversion section to the output of the code conversion and addition section. A code identification unit equipped with a pulse width check unit is connected to the output of a photoelectric conversion unit that receives an optical signal from the transmitter and converts it into an electrical signal, and a code identification unit equipped with a pulse width check unit is connected to the output of the code identification unit via a control unit. Connect the serial/parallel converter using
A match check in which the first data latch section is connected to the maximum 256 outputs of the parallel conversion section, and the code identification section is connected to the input/output terminal of each -th data latch section via the pulse number check and parity check section. A receiver is provided in which a maximum of 256 output points of the -th data latch section are connected to output stage latches connected to the latch control section, and the photoelectric conversion section and the latch control section of the transmitter are connected to each other. An optical multiplex transmission system characterized by connecting a photoelectric conversion section of a receiver with an optical fiber cable.