JPS5890851A - Detection system for collision of communication - Google Patents

Abstract

PURPOSE:To detect the collision of data, by encoding and transmitting signals transmitted to a data bus with a contention system so as to obtain one time slot for the maximum time width of a mark or a space. CONSTITUTION:Signals on a data bus are wave-shaped at a comparator 12 via an amplifier 12. The shaped signal is applied to a data terminal of a D type flip-flop 13 and also to a clock terminal of the flip-flop 13 via a leading detection circuit 14 comprising an inverter circuit 14a and an AND circuit 14D, and a delay circuit 15. The collision can be detected through an output of the flip- flop 13 to be ''1''.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention provides a system for reliably detecting the balance of data on a data bus when continuous type KL data communication is performed via the data bus. This paper relates to a communication center detection method that enables communication center detection.

[Technical background of the invention]

Recently, a system that connects multiple stages 17 via a data bus and mutually transmits necessary data between stations via the data bus has been attracting attention. Instead of a system in which the stage system is divided into 1 III divisions, or the use of paths is centrally managed to perform communication between stages,
The contention method, in which each stage controller detects the vacant state of a path and transmits data at a desired time, is attracting attention for reasons such as a simple system configuration.

However, in this content transmission method, when the data bus becomes vacant, multiple stages transmit data at the same time, which can cause data to collide on the bus and cause information to be lost. There is a risk that Therefore, various efforts have been made to promptly detect the collision of puri and stop data transmission.

What is shown in Fig. xi is an exclusive OR (IX- Oll
) Data bus 2 is compared in circuit 4 and based on the coincidence
This is an attempt to detect the collision between the signal transmitted from the own station and the signal from another station. However, the compliance element 5 compensates for the transmission delay of data (signal) and detects the collision between the signal transmitted from the own station and the signal from another station. This is to match the signal tie construction.

On the other hand, in the case shown in FIGS. 2(a) to 2(c), signals transmitted from multiple stages are connected to the data bus 2 as shown in FIGS. 2(a) and 2(b), respectively. k
When VA and VB exist, the level is as shown in the same figure (e
), paying attention to what occurs as shown in path line 2.
A collision is detected by discriminating the upper signal level using a predetermined threshold qt.

[Problems with background technology]

However, although the above-mentioned two types of collision detection methods can theoretically accurately detect a Kll collision, in reality, the signal transmitted via the data bus 2 is often accompanied by a large loss. The signal level difference caused by this loss is
In extreme cases, it can reach several 104B. In particular, in a transmission system in which a signal is transmitted via an optical fiber using light as a carrier, this problem occurs conspicuously. For this reason, the threshold value V! for signal level comparison for collision detection as described above! If given a constant value, the rise and fall characteristics of the transmission signal have a finite value in the well-known x5 0tx, which tends to lead to collision detection errors. The signal pulse width changes depending on the threshold value. For this reason, when detecting the dispersion of the signal pattern as shown in FIG. 111, there is a possibility that the change in the signal width may lead to erroneous detection of a collision.

[Invention OFA]

The present invention was made in consideration of the above circumstances, and the purpose of the present invention is to use the content-enabled KL method to quickly prevent collisions of transmitted data on the data bus. The object of the present invention is to provide a highly practical communication collision detection method that can reliably detect collisions.

[Summary of the invention]

The present invention uses the Contin VWy system to transmit signals to the data bus with the maximum time width of a mark or space t.
stBx becomes one time shift T. Encode and transmit, detect the transition point between the mark and space of this signal, and set the minimum time width of the mark or space of the signal to Tm1m.

Also, the maximum pulse width variation of the signal on the data bus 4τ
When the time above and smaller than l time slot T is ΔT, the time (tema!+
The point in time (TI!IILK-1
Data collision is detected by identifying the signal on the data bus at a point in time when Tm1m+ΔT) or more and at a time greater than or equal to time (Tm1m+ΔT). Taking advantage of the fact that the duration of a signal mark or space is not longer than a certain time Tmax, the time from the transition point between the mark and space (Tmg From the identification result of the signal at the time when T) is passed, it is determined whether or not an O signal exists from another stage, thereby reliably detecting a communication collision.
It becomes possible to tX As mentioned above, considering the minimum time width Tm1m of a mark or space, which is simultaneously determined by measuring the maximum time width 'f'aax of a mark or space for one time slot TK, in the following case,
Taking advantage of the fact that there are no signals in the range of time (Tall quantity 10) to (Twh * x - 1) from the transition point between mark and space, the signal on the data bus in the above time range is Therefore, overconfidence collisions can always be accurately detected without being influenced by the level-neutral rise and fall characteristics of the transmitted signal, and the advantage is enormous. Example] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. is the original signal,
(b) shows the Manchester signal.0 In this example, the original signal is ``1.0.1,'' where 1 time slot is i.
1,0,1. It shows a change of "O".

The Manchester signal generated to be sent to the data bus for this original signal is 11'' (mark), then the first half is ``O'' (space), and the second half is takes the form of "1" (mark), and when the above original signal becomes "0# (space)", the first half of one time slot is 11", and the second half is @0".
The Manchester signal takes the form of
Also, if x is the same as one time slot T,
The minimum time width Twini of the mark or space of the Manchester signal obtained by such encoding is defined as 1/2 of one time slot, that is, T/2. The maximum time width Tmax is specified, and a good signal is sent out in the Continuum V-type method from the multiple stages connected to the data bus and making up the transmission system, and data communication between each station is performed. In a system in which data communication is carried out in this manner, when a signal is sent from the station consisting of the station as shown in Figure 4 (Figure 4), and a signal is sent from another stage to the error shown in Figure 4 (b), the data is transmitted. Collision between the above two signals occurs on the bus, and the combined signal is as shown in the same figure (,). However, the signal sent on the data bus is )~
The signal pulse width fluctuation in the transmission system Kj11 caused by the 0-0 rise and fall times, which have finite time widths at the rise and fall times shown in (C), respectively, is the difference ΔT between the input pulse width and the output pulse width. The level of the signal sent out on the data bus should be kept constant at the original level, but in reality, it is inserted into the transmission line and suffers losses caused by couplers, etc. , it depends on the transmission distance.

However, if the signals on the data bus are discriminated and identified using a threshold value, then at the point in time when no collision occurs as shown in Fig. 14(d) K:1, the time lI of the mark or base of the transmitted signal is The specified time Tmax as
Considering the amount of variation Δτ, (Tmax + Δτ)
However, if a collision occurs, the time width will exceed (Tmax + Δt) due to an increase in the signal level or the temporal merging of the signals.In other words, as mentioned above, , since the maximum space time @ Tmaz of the space for the mark max of the sending signal is specified in one time slot T, the time Tmax is determined from the timing of the transmission signal.
When the next timing signal is detected, if there is no collision, both signals are marks and dangerous, or spaces, and it is impossible for them to cross each other 0 This ↓
If such a situation occurs, it can be said that a collision occurs between the signals from other stages on the data node and the signal is zero.

Therefore, in this method, the transition point, that is, the change point, between the mark and the space of the signal sent to the data node is detected, and from this timing force λ, the maximum time @ Tmax
The occurrence of a collision is detected by identifying whether the signal is a crab mark or a space.

In particular, considering the pulse width fluctuation amount Jτ, an auxiliary time JT of τ is set to identify the signal at the time (TnHx + ΔT) from the point of change between the mark and the space. In particular, it is designed to effectively and accurately detect data collisions on the data bus.

That is, as shown in FIG. 4 (), the transition point (rising point) of the identified original signal from the space to the mark on the data bus is detected, and the time (T + JT) is calculated from the point of detection.
At the timing shown in FIG. 4(f) after the
The signals on the data bus shown in d) are identified. If the signal at this timing is a space, it is assumed that no collision has occurred, and if the comparison signal is a mark, a collision has occurred because the mark continues for more than time (Tm4x + JT). Therefore, in this example, a collision is detected at timing te as shown in FIG. 4(g).

In addition, the signals sent to the bus in this way:
The minimum time 1ii Tmin (=2) is defined. In particular, the time width of a mark or space in the Manchester signal is one time slot) T (D
I/i.e. work or 1 time spot 2%
2 T. When no collision occurs, it is impossible for marks or spaces with an intermediate time width in the range of T/2 to T to be formed. Especially when considering the above-mentioned margin time ΔT tJJI, there is no mark or space in the range of time width (2-+pa) to (T-pa). If such a strange mark or space were to exist on the data bus, it could be said that it was caused by a collision with a signal from another station. Therefore, if we identify the signals on the path at the timings after the above times (A→・ΔT) and (T-PA) from the transition point between mark and space, we can detect communication collision from the identification results. becomes possible.

As is clear from the above explanation, since the signal sent to the data bus has the characteristic that one time slot T is the maximum time @Tmax of a mark or space, this is the point of change between a mark and a space.

The data bus at the time when (TmBx+Δd) has passed.

To identify the above signal, it is necessary to identify the signal on the data bus at each time point t or the time (71111m + 4 units) s (Tmax - 1 unit) from the above change point. Therefore, it is possible to easily and accurately detect the center of the O signal on the data bus. Even if the signal synthesized by each collision shows a rough level change as shown in Figure 4(,), the signal level from the nine stages is attenuated. Even if this is the case, if signal identification is performed with a threshold value low enough to identify the signal sent from the own stage ring, it is possible to accurately detect a collision based on the above-mentioned time relationship.

Therefore, communication collisions on the data bus in which data is transmitted using the Continuum V-wire method can be easily and accurately detected, which has great practical advantages.

FIG. 5 is a diagram showing an example of the configuration of a collision detection circuit that performs collision detection according to the method of the present invention described above.

A signal on the data bus is inputted via an amplifier 11, and then discriminated by a predetermined threshold value and waveform-shaped by a comparator 12. This signal is supplied to the data terminal of the D-lid flip-flop 13 and is also input to the rising edge detection circuit 14.

This rising edge detection circuit 14 detects, for example, the input signal in the first
and input the above input signal to the inverter circuit 1 terminal.
AND circuit 14b which inputs to the second terminal a signal which is inverted via 4a and delayed by the inversion processing time.
Then, when the previous 11tlt changes from "0" (space) to 11" (mark), the AND circuit 14b receives the "l# flaw signal to the first terminal, and then, The signal supplied to the 20th terminal changes from 11# to "0"K with a slight delay. This slight time delay is used to detect the rising edge of the signal. Then, the pulse signal obtained by this rising edge detection is delayed by a time (TmB x 10) through a delay circuit 115 composed of monostable multivibrators, and then used as a clock signal for the flip-flop 13. Given 0 and 1, the flip-flop 13 is
The signal on the path is identified and input at a timing (Tmax 10) after the rising edge detection timing. Therefore, by monitoring the output Q of this tree flop 13, when the output Q becomes "1", it is possible to detect the communication center, assuming that a mark signal is present at a timing that cannot normally exist. The collision will be detected by identifying the signal on the bus at the time of passing the time (TazB + ΔT) from the change point of the mark 〆\ from the space. Also, Figure 6 shows the time width. This shows an example of the configuration of a detection circuit that detects a collision by utilizing the absence of a mark or space (T-T). D[Flip-flop J#, IF is used to detect the rising edge detector 1
The output of t7Ij is delayed by a time (T10) through a delay circuit 18 and used as a clock for the tree flop 16, and the output of the rising edge detector 14 is delayed by a time (T-1) through a delay circuit 1g. and is configured to provide each clock as a clock to the tree flop 1r. Therefore, the mode of the signals detected by these flip-flops 16.11 is "
If 0# is set, ot7t is the time when no collision occurs, and either one of the tree flops Jl
The next time 11m is set in l, JP is when a collision has clearly occurred.

When t + 11" is set in both double flip-flops 16IF, there are cases where a collision occurs and cases where the time width of the mark corresponds to the time slot T. In this case, it is certain that there is no collision. Therefore, the outputs of the flip-flop 1g and IF are passed through an exclusive OR (xi-OR) circuit 2o to obtain
It becomes possible to detect collisions. However, 1 As mentioned above, the flip-flop is 1g.

12 are both set to "1", no collision detection signal is obtained as the output of the ΣX-OR circuit 20 because it is uncertain whether there is a collision or normality.

The detection circuit shown in Figure 7 is a combination of the detection circuits shown in Figure 5 and Figure 6, and the output of the flip-flop 13 and Σ
Obtaining the output of the X-OR circuit 2o through the OR circuit z1 requires the following times:
A collision detection signal is obtained from the signal identification result at a certain point in time. 1 for doing it this way, even more! It becomes possible to perform highly accurate collision detection.

As mentioned above, the II center detection according to the method of the present invention is very simply constructed and can be performed in one circuit in nine circuits. Moreover,
Even if the level oii and tsuta signals from multiple stations collide and the signal level is determined based on the above-mentioned temporal relationship, reliable detection can be performed, so the effect is enormous. .

Note that the present invention is not limited to the above embodiments. In the embodiment, a Manchester signal will be explained as an example, but the present invention can also be applied to, for example, a DMI encoded signal, as long as the maximum duration of a mark or space is defined in one time slot. It goes without saying that collision detection can be performed in a similar manner not only by focusing on marks but also by focusing on spaces. In short, the present invention can be implemented with various modifications without departing from the gist thereof.

[Brief explanation of drawings]

Figure 1 is a diagram showing an example of a conventional collision detection circuit, #I2 (&) to (C) are diagrams for explaining collision detection by signal collision and level determination, Figure 3 (a) # (b) shows the transmission signal form according to the present invention, Figures 4-) to (g) are signal diagrams showing the collision detection function of an embodiment of the method of the present invention, and Figures 5 to 7 are respectively FIG. 2 is a diagram showing an example of the configuration of an S center detection circuit according to the method of the present invention. 11...Amplifier, 12...Comparator, 13.18°1
9... D-type flip-flop, 14... Rise detection circuit, 11.1-, IF... Delay circuit, go... Exclusive OR circuit, 21... OR circuit. Figure 1 Figure 2 Figure 3 Figure 4-

Claims (2)

[Claims] (1) Maximum mark or space time qTwan
means for detecting the transition point between the mark and space of the signal sent to the data bus by specifying one time slot TK for g, and the mark t or space defined by the maximum time width Tmax (D specified KL) The minimum time for - Tm
When i m is the time that is greater than or equal to the maximum pulse width variation JT of the signal on the data bus and less than the l time slot T, then the time ('rmax+JT) has passed from the transition point between the mark and the space. 9 time points, or less than or equal to time (Tmax-JT) and time (Tmi
m+JT) Overconfidence** detection method 0 characterized by comprising means for identifying the signal on the data bus at the above point in time and detecting the compensation center. (2) The means for specifying the maximum time g Tmax of the space of the signal 0-2-t sent to the data bus to one time slot is to encode the original signal by Manchester encoding or DMI encoding. , the minimum time width Twin of the mark or space of the encoded signal is l
1/ of time slot T! A communication center detection method according to claim 1, which is defined in claim 1.