JPH04196642A - Method of controlling communication network - Google Patents

Abstract

PURPOSE:To prevent the decrease of a through-put and the increase of a through-put delay by performing respectively a specific processing at plural node devices and plural transmission/reception terminals connected to a tree-like hierarchical structure. CONSTITUTION:A network is made as the tree-like hierarchical structure, the transmission/reception terminals 2 is connected to the lowest node device 1c of the node device 1. Also, a first arrival identification signal is outputted by a first processing, the controls of the normal connection of a lower direction port to an upper direction port and the normal connection of a first arrival identification signal input port to a first arrival identification signal output port is attained by a second processing and the control of the highest node device 1a is attained by outputting the first arrival identification signal from the first arrival identification signal output port by a third processing. Further, the detection of communication failure by detecting the first arrival identification signal by the transmission/reception terminal 2 or by detecting that by the node device 1 is attained by the fourth processing. Thus, the high through- put and the excellent through-put delay characteristics are obtd.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a communication network control method suitable for controlling the O8I physical layer and network layer of a LAN (Local Area Network).

2. Description of the Related Art In general, in a LAN, a plurality of stations access a single transmission path for communication, so there are various access methods.

For example, C3MA/CD (carrier detection multiple access/
There is a LAN using the collision calibration method. This is IEEE
It is a LAN such as Ethernet or Star Run defined by 802.3, and is characterized by a simple protocol. In this method, the terminal starts transmission after confirming the carrier on the transmission path.

When a collision occurs, it stops transmitting and enters backoff.

This method has low delay and high efficiency when traffic is low, but when traffic becomes high, collisions occur frequently and throughput is limited. Then, as the traffic becomes higher, the delay increases and the throughput decreases.

Therefore, a much higher data rate than the traffic needs to be used, resulting in higher equipment and cable costs.

There is also a token-based LAN. This is IEEE
Token bus specified by 802.4, IEE
L of token ring etc. specified in E 802.5
It is AN. In this method, terminals obtain transmission rights in order and transmit. Compared to the C3MA/CD method, this method has less delay during high traffic, has fewer throughput limitations, and is more efficient. However, the delay does not decrease much even during low traffic, and
The equipment cost is high because the protocol is complicated.

In addition, the cost of equipment becomes extremely high.
Nor can high data rates be used.

Further, as one of the C3MA/CD systems, there is one disclosed in Japanese Patent Publication No. 58-40384, for example. According to this method, even if a collision occurs, packets are not destroyed and a single communication is established, resulting in less increase in delay when traffic is high and high efficiency. However, since collision determination is performed by comparing the addresses of output packets and input packets, the equipment cost is high and IEEE 802
．． It also deviates from the standard of 3. Therefore, it is necessary to add a new full-duplex function. Versatility and compatibility are important elements of a communication system, and failure to meet standards will reduce the marketability of the device.

For this reason, the present applicant has made a proposal to improve the access method in Japanese Patent Laid-Open No. 192242/1999. According to this publication, node devices are connected in a tree, and when there is an input to a lower direction port, the node device gives priority to the first input port, and by giving priority to the input of the upper direction port, one communication is guaranteed. We are trying to make this happen. Collision inspection is a method of comparing the presence or absence of signals at input and output ports, and the network interface device satisfies the IEEE 802.3 standard. Therefore, the network interface device does not become highly rigid unlike the system disclosed in Japanese Patent Publication No. 58-40384. Furthermore, there is no reduction in throughput characteristics during high traffic, which is advantageous over C3MA/CD type Ethernet.

Problems to be Solved by the Invention However, in the method proposed by the present applicant, when switching the input in the lower direction to the input in the upper direction and outputting it to the terminal, IEEE 802 is added to the preamble part of the output signal.
．． Codes rl and IJ, which correspond to FCDs of three standards, may occur, resulting in reading errors at the terminal. Also, since a long "1" signal is used, AC coupling (R8-422)
cannot be used.

Therefore, there is a need for a communication network control method that uses the simple protocol of the existing IEEE 802.3 standard, allows one communication to be established even when a collision occurs, and does not reduce throughput even during high traffic.

Therefore, the applicant has proposed a method disclosed in Japanese Patent Laid-Open No. 1-192242.

These points will be further explained.

First, Figure 10 shows the throughput delay (T), which is well known for evaluating the performance of bus-type LANs.
In other words, the amount of communication transmitted versus the waiting time until communication is established is shown. Here, throughput and delay characteristics are shown for typical LAN systems, the C3MA/CD system and the token ring system.

Here, as described above, the C3MA/CD method has a smaller delay than the token ring method when the traffic is low, but the delay is large when the traffic is high. As traffic increases further, throughput decreases. This is because in the C3MA/CD system, as the traffic increases, the probability of collision (packet collision) increases and the delay increases due to retry (packet retransmission). Furthermore, since packets destroyed by collision occupy the network meaninglessly, the capacity of the network decreases accordingly. Therefore, when the traffic becomes too high to ignore the capacity of the network, the throughput decreases.

On the other hand, according to the method of Japanese Patent Application Publication No. 58-40384 mentioned above, even if a collision occurs, one communication is established. has been remodeled. This prevents the delay from increasing due to retry. In addition, packets destroyed by collisions do not occupy the network meaninglessly.
This also prevents throughput from decreasing. In these proposed examples, the network has a tree-like hierarchical structure, and the first-arriving packet is validated at the top node device and broadcasted throughout the network, and other packets are rejected as invalid. is prevented from being destroyed by collision.

The state of a packet that is rejected as invalid will hereinafter be referred to as "communication failure."

However, the terminal of the Japanese Patent Publication No. 58-40384 system is
To determine whether or not a packet sent by oneself has failed, the packet being transmitted is compared with the received packet sent back from the highest node device, and if they are different, it is determined that communication has failed. Therefore, CSMA
Although it is a /CD system, it is different from that specified by IEEE 802.3. Therefore, the NIU (network interface: a communication board provided in the terminal) needs to be newly designed, and it becomes expensive because it requires a full-duplex communication function.

On the other hand, according to the method disclosed in Japanese Unexamined Patent Publication No. 1-1.92242, failure of communication is detected without any concept of address. Therefore, the communication failure detection device can be placed inside the node device, and basically the NIU and protocol of the terminal can be used without modification. However,
The method disclosed in Japanese Patent Application Laid-Open No. 1-192242 has the following problems, and there are restrictions on the use of the protocol defined in IE and EE 802.3 and the NIU without modification. First, the time required for the preamble length must be significantly larger than the network time constant. In the Ethernet example, the preamble length is 3.2 μsec
The network length (twice the distance from the top node device to the farthest bottom node device or terminal) is 640 m (taking into account optical communication, 1μ5ec
−+200 m) or less. Second, some terminal NIU transceivers are of a type in which synchronization cannot be immediately achieved when the preamble changes midway and the received signal loses synchronization midway.

Regarding this point, an improved technique is disclosed in Japanese Patent Application Laid-Open No. 192242/1992, but this is not applicable to all commercially available NIUs.
There is no guarantee that the method can be applied without problems to other transceivers.

Means for Solving the Problems The invention according to claim 1 has a single upper direction port and a plurality of lower direction ports, and connects the first lower direction input port to the lower direction output port by first-come-first-served logic. ,
A single node device with the function of connecting an upper direction input port to a lower direction output port or a plurality of node devices connected in a tree-like hierarchical structure via a transmission path is provided to detect the presence or absence of a signal at an input port. a plurality of transmitting/receiving terminals connected to the node device and having a function of allowing transmission only when there is no signal and retransmitting when a signal collision occurs;
Of the signals transmitted almost simultaneously from multiple transmitting and receiving terminals,
In a communication network control method in which only the first-arriving signal from the highest-level node device is broadcast into the communication network and other signals are excluded, connections between the node devices include a connection from a lower-level node device to a higher-level node device. Using at least three signal lines for transmitting a signal, a signal from an upper node device to a lower node device, and a first-arrival identification signal from an upper node device to a lower node device, The structure has a direction input port, a first arrival identification signal input port, and a lower direction output port, and the lower direction port has a lower direction input port, a first arrival identification signal output port, and a lower direction output port. When there is a signal at any of the input ports, the node device connects the first input port in the lower direction to the output port in the upper direction, disconnects the other input ports in the lower direction, and outputs the first arrival identification signal. A first-arrival identification signal is output from the port, and secondly, when there is a signal at the upper direction input port, the node device connects this to all lower direction output ports regardless of the presence or absence of a signal at the lower direction input port. Also, when there is a signal at the first-arrival identification signal input port, this first-arrival identification signal input port is connected to the first-arrival identification signal output port, and thirdly, when there is a signal at any of the lower direction input ports, the top node The device detects the first input port in the lower direction, disconnects the other input ports in the lower direction, connects the first input port in the lower direction to all the output ports in the lower direction, and connects the first input port in the lower direction to the first arrival identification signal output port. Fourthly, the transmitting terminal outputs a first-arriving identification signal from the first-arriving identification signal input port, so that the signal outputted by the transmitting terminal does not become the first-arriving node of the highest node device.
When it detects that communication has failed, it controls retransmission after a predetermined time, or the node device to which the terminal is connected uses the first-arrival identification signal of the first-arrival identification signal input port to transmit the signal output from itself to the highest node. When it is detected that the device is not the first to arrive and communication is not established, it outputs a communication failure signal from the lower direction output port corresponding to the input port that had the signal including the first arrival input port among the lower direction input ports. , If this communication failure signal is present at the input port, the transmitting terminal determines that communication is unsuccessful and performs retransmission control after a predetermined period of time.Fifth, the node device determines whether or not there is a signal at the lower direction input port and the upper direction input port. The end of communication is detected by detecting the loss of signals from both input ports, and the connection state of each port is initialized.

In addition to the invention set forth in claim 1, in the invention set forth in claim 2, the first-arrival identification signal indicates that the signal outputted by itself is a delayed input in the higher-order node device; When there is a signal on any of the direction input ports, the node device connects that first input port to the upper direction output port, disconnects the other lower direction input ports, and responds to the lower direction first input port. A first-arrival identification signal is output from a first-arrival identification signal output port that is not configured, and secondly, the uppermost node device connects the upper direction output port back to the upper direction input port to perform common control with other node devices; When there is a signal on any of the lower direction input ports, that lower direction first input port is connected to the upper direction output port, the other lower direction input ports are disconnected, and the first arrival identification that does not correspond to the lower direction input port is performed. By outputting the first-come-first-served identification signal from the signal output port, if there is a return signal at the upper direction input port, it is output from the lower direction output port and the third
When there is a signal at multiple lower direction input ports among the lower direction output ports to which the transmitting and receiving terminal is connected, the node device to which the transmitting/receiving terminal is connected responds to the lower direction input port with the input signal with a delay. If a communication failure signal is output from the lower direction output port, or if there is a first-arrival identification signal at the first-arrival identification signal input port, the communication is output from the lower-direction output port corresponding to the lower-direction input port where the signal was located, including the lower-direction first-arrival input port. A communication failure signal is now output.

In the invention according to claim 3, a communication network failure detection unit is provided in the communication network connection device of the transmitting and receiving terminal, and the failure detection unit detects the input to the input port within a predetermined communication network time constant starting from the start of transmission. When it detects that there is no signal, it is determined that the communication network has failed.

In the invention according to claim 4, a failure detection section is provided in at least the node device to which the transmitting/receiving terminal is connected;
If this failure detection unit detects that there is no input signal at the upper direction input port within a predetermined communication network time constant starting from the detection of the signal at the lower direction first input port, it determines that there is a failure in the communication network, and the second In addition, through this failure inspection, the transmitting/receiving terminals connected to the lower direction output ports corresponding to the lower direction first-come-first-served input ports or all the lower direction ratio ports are notified of the failure detection.

In the invention set forth in claim 5, when the failure detection unit detects a failure, it outputs a communication failure signal from the lower direction output port or all lower direction ratio ports corresponding to the lower direction first-come-first-served input port connected to the transmitting/receiving terminal. I did it like that.

In the invention as set forth in claim 6, firstly, the node device to which the transmitting/receiving terminal is connected connects the lower direction input port to which the transmitting/receiving terminal is connected back to the corresponding lower direction output port; , if there is a signal at the upper direction input port and a first arrival identification signal at the first arrival identification signal input port, the return connection of all the lower direction input ports to which the transmitting/receiving terminal is connected to the corresponding lower direction output port is cut off. , connect the upper direction input port to this lower direction output port, and thirdly, if there is no first arrival identification signal at the first arrival identification signal input port and there is a signal at the upper direction input port, transmission/reception from other than the lower direction first input port is performed. The loopback connection of the lower direction input port to which the terminal is connected to the corresponding lower direction output port is cut off, and the upper direction input port is connected to its lower direction output port.

In the invention set forth in claim 7, when the failure detection unit of the node device detects a failure, it corresponds to that of the lower direction first-come-first-served input port to which the transmitting/receiving terminal is connected or all the lower direction input ports to which the transmitting/receiving terminal is connected. The loopback connection to the lower direction output port is now disconnected.

In the invention described in claim 8, a communication network monitoring device having a higher priority than other transmitting/receiving terminals is connected to the upper direction port of the highest node device, and first, the highest node device internally The upper direction output port is connected back to at least the upper direction input port, and secondly, when the upper direction input port receives an input signal from the communication network monitoring device,
The highest level node device disconnects the return connection and connects the upper direction input port to all the lower direction output ports.

According to the invention set forth in claim 1, the first process outputs the first-arrival identification signal, and the second process constantly connects the lower direction port to the upper direction output port, and the first-arrival identification signal input port outputs the first-arrival identification signal. The constant connection to the identification signal output port is controlled, the third process controls the top node device by outputting the first-arrival identification signal from the first-arrival identification signal output port, and the fourth process controls the first-arrival connection by the transmitting and receiving terminal. Communication failure is detected by identification signal detection,
Alternatively, failure of communication is detected by first-arrival identification signal detection by the node device, and inspection control for termination of communication is performed by the fifth process. Therefore, while using the existing IEEE 802.3 standard protocol, It is possible to obtain higher throughput and excellent throughput/delay characteristics compared to the conventional method. In particular, according to the present invention, in consideration of compatibility with the IEEE 802.3 standards such as Ethernet IOT and Star Run 10, which are currently being standardized, the connection between node devices is twisted.
By using pairs and utilizing one of the two unused pairs as a first-come-first-served identification signal line, the configuration of the node device becomes simple. In addition, the preamble of the message packet is no longer deleted by the first-come-first-served identification signal.
There is no need to carefully set the first-come-first-served identification signal.

Here, there are several modifications to the existing IEEE 802.3 standard protocols and network interface devices, and even for these modifications, the network interface device of the transmitting/receiving terminal can be modified according to the invention described in claims 2 to 8. The signals input to the system are almost the same as those of the IEEE802.3 standard, and can be used without modification.

That is, according to the invention as claimed in claim 2, the first-arrival identification signal is given a meaning, the first process identifies the lower direction output port that outputs the first-arrival identification signal, and the second process identifies the first-arrival identification signal. is immediately output and the control of the highest node device is shared, and the third process immediately outputs a communication failure signal.

According to the third aspect of the present invention, in the configuration including the failure detection section, a communication network failure is determined by the transmitting/receiving terminal.

According to the fourth aspect of the invention, in a configuration in which a node device includes a failure detection section, the first process detects a failure in the communication network in the node device, and the second process communicates the failure detection. .

According to the invention set forth in claim 5, the communication failure signal is used as a substitute for transmitting the failure detection.

According to the sixth aspect of the invention, the input of the transmitting/receiving terminal connection port is returned to the output port and the return is canceled.

According to the seventh aspect of the present invention, failure detection is transmitted by disconnecting the input port from the output port instead.

According to the eighth aspect of the invention, the communication network monitoring device is connected to the highest node device.

Embodiment An embodiment of the present invention will be explained based on FIGS. 1 to 9.

First, in this embodiment, the communication network time constant (network time constant), which is one fixed period of time, is the time starting from the detection of the signal of the lower direction first-arrival input port, and is the time that starts from the detection of the signal of the lower direction first-come-first-served input port. It is defined as a propagation delay time twice the distance to a lower node device or a transmitting/receiving terminal, and is normally used with an additional margin time required for signal inspection, etc.

Furthermore, although the following three predetermined signals are used, they are not limited to this example as long as they are distinguishable.

First, the first predetermined signal is a first-arrival identification signal. This signal is a signal that informs the node equipment to which the sending terminal is connected whether or not the transmitted message packet has become the first-come, first-served input of the highest-level node equipment. I conclude that it was not. This first-arrival identification signal may be any signal, and if the connection between the node devices is AC coupling, an appropriate signal (for example, a clock signal) that can pass through a transformer is used.

The second predetermined signal is a communication failure signal. When using the Ethernet protocol and the NIU, a 10 MHz signal is ultimately output to the collision wire of the transceiver cable. Also, the Star Run protocol and NIU
If this is the case, the jamming signal will eventually be output to the receive line of the transceiver cable.

The third predetermined signal is a fault signal. This can be done by adding a failure signal line to the transceiver cable, or by using the communication failure signal described above instead, or by looping back the lower direction input port to which the node device is connected to the lower direction output port. This can be replaced by disconnecting the connection.

Next, FIG. 2 shows the basic configuration of the network of this embodiment. As shown in the figure, the network has a tree-like hierarchical structure, and the node device 1 is the highest-level node device 1a.
, an intermediate node device 1b, and a lowest node device IC. These node devices 1a, lb, and lc are all the same, and the transmitting/receiving terminal 2 can be connected to any of the node devices 1. However, it is also possible to configure the network with only a single node device 1. Here, connection between the node devices 1 is performed by an upper direction signal line 1d, a lower direction signal line 1e, and a first arrival identification signal line if. In this example, the Ethernet NI
The connection between the lower port of the node device I and the terminal 2 is the transmit line 2a (as seen from the terminal 2 side), the receive line 2b (as seen from the terminal 2 side), and the collision line. This is done by the pre-sense line 2C.

The basic operation of such a network is shown in FIGS. 3 to 7. FIG. 3 shows step 1, in which a certain terminal 200 sends a message packet 200a to the terminal 200.
4, and at the same time, the other terminals 201, 20
2,205 are message packets 201a, 20, respectively.
2a and 205a are started to be transmitted. In contrast, the node devices 100, 101, 102, 104, 10
5 are message packets 200a and 202a, respectively.
, 205a, 200a, and 205a are the first-arrival input signals of the lower-order ports, so each is output from the upper-order port. Node devices 100, 101.
102 are message packets 200a.

202a, 201a, 202a, 205a from the lower direction port to the terminal 200, 201.202.205
Output to. Further, the node device 100 whose terminal is connected to the lower direction port outputs a communication failure signal from the collision line 2C of the lower direction port to which the input is delayed. On the other hand, the node devices 104 and 105 to which node devices are connected to lower direction ports output first-arrival identification signals from lower-order ports other than the first-arrival input ports.

FIG. 4 shows step 2. Since the message packet 200a is the first-arrival input signal of the lower-order port, the top-level node device 106 returns it and outputs it from all the lower-order ports, A first-come-first-served identification signal is output from the identification signal output port. In addition, in step 1 shown in FIG. 3, the node device 101 has a first-arrival identification signal in the first-arrival identification signal input port and an input in the lower direction input port, so the collision line of the lower direction port where the input was received is A communication failure signal is output from 2c. Since the terminal 201 received a communication failure signal on the collision line 2C in step 1, it stops outputting.

FIG. 5 shows step 3, in which the node device 1o4゜105
is the message packet 2 input to the upper direction port
00a is output from all lower directions.

Further, the node device 105 outputs the first-arrival identification signal input to the first-arrival identification signal input port from all the first-arrival identification signal output ports. On the other hand, since there was a communication failure signal on the collision line 2c in step 2, the terminal 202 stops outputting.

FIG. 6 shows step 4, in which the middle layer node device 10
0, 101, 102, and 103 output the message packet 200a input to the upper direction port from all the lower direction. Further, since the first-arrival identification signal was present in the upper direction port in step 3, the node device 102 outputs a communication failure signal from the lower direction collision/presence line 2c to which the input was received.

FIG. 7 shows step 5, in which the terminal 205 stops outputting since there was a communication failure signal on the collision line 2C in step 4.

In this way, message packet 200a reaches terminal 204 and all other terminals 2.

FIG. 8 shows the basic configuration of the node device 1 that operates as described above. The node device I has an upper direction module/○ module (U-I/OModule) 3 and a lower direction I10 module (D-IloModule) 4
and a connection control device (Kernel
) 5, and between the module 3.4 and the connection control device 5 are an upper direction input channel (UIC) 6a, an upper direction output channel (UQC) 6b, a lower direction input channel (DIC) 6c, and a lower direction output channel. (DOC)
6d and a terminal connection control channel (TCC) 6e.

The upper direction input channel 6a includes a set of upper direction signal input channel (iS) 6aa and a first arrival identification signal input channel (
irS) 6ab, and an upper direction output channel (U
OC) 6b is a set of upper direction signal output channels (oS)
Consisting of 6ba.

The lower direction input channel 6C consists of four sets of lower direction signal input channels (ic)-i3) 6ca, and the lower direction output channel 6d consists of four sets of lower direction signal input channels (0
0~o3) 6da and first-come-first-served identification signal input channel (oro
-or3) Consists of 6db.

Furthermore, the terminal connection control channel 6e includes four sets of output control channels (eo to e3) 6ea and collision presence channels (cO-c3) 6eb.

The configuration of the upper direction I10 module 3 and the lower direction I10 module 4 is shown in FIG. The upper direction I10 module 3 includes an upper direction node connection module 3a, a signal return module 3b, and a network monitoring device connection ■/○ module 3c, and the module 3a connects the intermediate node device 1b and the lowest node device IC. The modules 3b and 3c are used in the top node device 1a. Further, the lower direction I10 module 4 includes a lower direction pair node connection I10 module 4a and an opposite end unconnected I10 module 4a.
There are 10 modules 4b, which are used in all node devices 1.

The upper direction node-to-node connection ■/○ module 3a is composed of an R3422 transmitter 3aa, an R3422 receiver 3ab, and a transformer 3ac for AC coupling when twisted pair wires are used. Further, the signal folding module 3b simply connects the upper direction output channel 6b to the upper direction input channel 6a.

When the network monitoring device connection I10 module 3C receives an input signal to the upper direction port 3ca, it disconnects the upper direction output channel 6b from the upper direction human power channel 6a, and connects the upper direction port 3ca to the upper direction input channel 6a. Connecting. When there is no input signal to the upper direction input port 3ca, the upper direction output channel 6b is turned on again.
is connected back to the upper direction input channel 6a, and
The upper direction input channel 6a is connected to the upper direction input port (oM
) 3cc and a logic circuit group 3cf. Further, the network monitoring device connection I10 module 3C may be provided with a first-arrival identification signal input port (lrM) 3cc as required. As a result, the network monitoring device connected to the upper direction port 3ca of the network monitoring device connection I10 module 3C can give the highest priority interrupt even when the network is communicating.

On the other hand, the lower direction node-to-node connection I10 module 4a is
When using twisted pair wires, it consists of an R3422 transmitter 4aa, an R5422 seam 4ab, and a transformer 4aC for AC coupling.

Terminal connection/○ Module 4b is an example of a configuration compatible with an Ethernet transceiver cable. That is, an R3422 transmitter 4ba and an R3422 receiver 4bb for physically mutually converting the signals of the lower direction input channel 6C1, the lower direction output channel 6d, and the terminal connection control channel 6e and the transceiver cable signal, and an R3422 receiver 4bb for AC coupling. It consists of a transformer 4bc, an AND gate 4bd for converting the signal of the collision presence channel 6eb of the terminal connection control channel 6e into an Ethernet collision presence signal, and a clock generator 4be. In Ethernet, the status of terminals scheduled for communication (with which a session is established) is monitored at any time, and when a collision presence signal is detected, it is determined that communication has not been established. Therefore, it is appropriate to convert the communication failure signal into a collision presence signal.

In addition, the pressure control channel of the terminal connection control channel 6e (
eo to e3) 6ea controls the return connection of the lower direction input port 4bg to the lower direction output port 4h and the connection control of the lower direction signal output channel (00 to o'3) 6da to the lower direction output port 4bh. It consists of a logic circuit group 4bf for. The control of the pressure control channel 6ea of the terminal connection control channel 6e has the following three functions.

■ Lower direction output port 4 of lower direction input port RO bg
Return to bh and lower direction signal output channel 6da
Connection control to the lower direction output port 4bh.

■ Message packet (message packet 200 in the illustrated example) that is first input to the top node device 1a.
In the node device (node device 100 in the illustrated example) to which the transmitting terminal (terminal 200 in the illustrated example) of a) is connected, the lower direction output channel 6d of the lower direction output channel 6d to the lower direction output port to which this terminal is connected. By disconnecting the signal output channel 6da, message packets returned by the highest node device 1a are prevented from being input to this terminal (node devices to which the transmitting terminal is already connected can prevent message packets from being input to this terminal). ).

In other words, the exception connection control is as follows.

■ A network failure is communicated to the transmitting terminal by disconnecting the loopback connection of the lower direction port to which the transmitting terminal is connected (exception connection control in ■).

In Ethernet, the transmitting terminal monitors the signal of the input port, so eliminating the input signal by disconnecting the upper direction port to which the transmitting terminal is connected will signal a network failure. This is an appropriate method for transmitting information to the terminal.

Next, the basic configuration of the connection control device 5 is shown in FIG. Basically, the upper direction input signal detection unit (USDU
) 7, a failure detection unit (DDU) 9, and an end detection unit (ED
U) 10 and lower direction input signal detection unit (DSDU) 11
, a first-arrival input port detection unit (FDU) 12, a communication failure signal generation unit (CPGU) 13, and a first-arrival identification signal generation unit (
“HNGU) 14 and switching gate unit (SGU
) 15 and an output control unit (OIEU) 16.

The upper direction input signal detection section 7 includes a flip-flop 7a for detecting the signal of the upper direction input signal channel and a flip-flop 7b for detecting the signal of the first arrival identification signal input channel.

The failure detection unit 9 detects that the time from the lower direction input signal to the upper direction input signal exceeds the network time constant.
Shift register 9 for detecting network failures
It is composed of a.

As long as the network does not fail, the input signal is guaranteed to be present within the network time constant.
Failure detection using this method is reasonable.

The end detection unit IO is composed of a shift register 10a for detecting the end of communication when both the lower direction input signal and the upper direction input signal disappear, and this output initializes the node device 1.

The upper direction input signal detection section 11 includes a flip-flop group 11a for detecting the presence or absence of an input signal.

The first-arrival input port detection unit 12 includes a flip-flop group 12a for detecting the presence or absence of an input signal, a NAND gate group 12b for prioritizing, and a NAND gate group 12b for prohibiting the operation of the flip-flop group 12a. It consists of a gate 12c, an inverter 12d, and a NAND gate group 12 group 2e for controlling the output of the NAND gate group 12b.

The communication failure signal generation unit 13 includes a communication failure signal generation circuit (
○PGE-0 to 3) l3a, when the upper direction input signal detection section 7 detects the first arrival identification signal, the lower direction input signal detection section 11 sends a communication failure signal to the channel where the input signal was detected. Output. Also, when there are multiple inputs to the lower direction input port, a communication failure signal is output to the channel that received the delayed input according to the control signals of the lower direction input signal detection section 11 and the first-arrival input port detection section 12. .

The first-arrival identification signal generation section 14 is an AN for generating a first-arrival identification signal based on the control signal from the first-arrival input port detection section 12.
It consists of a D gate group 14a.

The switching gate section 15 is the first input port detection section 1
NAND gate group 15 upper group 5a for selecting one from the lower direction input channels 6c and connecting it to the upper direction output channel 6b by the control signal from 2, the upper direction input channel 6a and the first-come-first-served identification signal. It consists of an OR gate group 15c for outputting the signal from the generating section 14 to the lower direction output channel d, and a connection for outputting the signal of the upper direction input channel 6a to the lower direction output channel 6d.

In addition, the output control section 16 is an output control pre-stage circuit (OEPE).
16a and output control circuits (OEE-1 to OEE-3) 16b. The return connection of the unconnected I10 module 4b is cut off, and the lower direction output channel 6d is connected to this output port. Furthermore, the upper direction input signal detection section 7
When detects the input of the upper direction signal input channel, control signals from the lower direction input signal detection section 12, first arrival input port detection section 12, and first arrival identification signal generation 114 detect the unconnected I10 at the opposite end for channels other than the first arrival input port. The output port of module 4b is disconnected and the lower direction output channel 6d is connected to this output port.

All such control of the output control unit 16 is performed via the terminal connection control channel 6e to the unconnected r10 module 4b at the opposite end.
is carried out against

For detection of the first-arrival identification signal at the terminal, the first-arrival identification signal line 1 used for the connection between the node devices 1 shown in FIG. 2 may be used between the node device 1 and the terminal 2. That is, the terminal only needs to detect the presence or absence of a signal on the first-arrival identification signal line if.

Further, network failure detection in a terminal may be realized by adding the same configuration as the failure detection unit 9 shown in FIG. 1 to the NIU of the terminal device.

A communication failure signal may be used as a method for transmitting network failure detection to the terminal. In Ethernet, if retransmission (retry) continues for a certain number of times or more, retransmission is stopped (retry out). Since this corresponds to when an abnormality occurs in the network, it is appropriate to output a communication failure signal.

The network monitoring device may be realized by connecting the same device as the terminal 2 to the network monitoring device connection I10 module 3c shown in FIG. The collision presence signal and carrier detect signal (input port signal detection) may be ignored and output, or communication may be forcibly interrupted by outputting a first-arrival identification signal. .

Further, regarding the first-arrival identification signal, each node device may output the first-arrival identification signal from a lower-order output port that does not correspond to the lower-order first-arrival input port. As a modification, the first-arrival identification signal may be output from the lower direction output port corresponding to the first-arrival input port. According to this modification, the configuration is such that the first-arrival identification signal arrives only at the node device connected to the transmitting terminal of the message packet that is the first-arrival input of the highest-level node device. Such a configuration can be easily realized by simply changing the upper direction input signal detection section 7.

Effects of the Invention Since the present invention is configured as described above, even if a collision occurs, a packet is not destroyed and a single communication is established.
It is highly efficient with less increase in delay when traffic is high, and communication establishment/failure is immediately detected including collisions, so processing efficiency is high when communication is not established.Furthermore, it has a communication network failure detection function. With the addition of , it is also possible to detect failures in the communication network, and in particular, considering the compatibility with the IEEE 802.3 standards such as Ethernet 10T and Starlan 10, which are currently being standardized, it is possible to detect failures in the communication network. Since twisted pairs are used and one of the two unused pairs is used as a first-come-first-served identification signal line, the configuration of the node device can be simplified, and the preamplifier for message packets can be simplified. There is no need to carefully set the first-arrival identification signal, and since the node equipment detects the establishment/failure of communication and failure of the communication network, Terminals having the existing modified IEEE 802.3 standard protocol and network interface device can be connected convertibly without modification.

[Brief explanation of the drawing]

1 to 9 show an embodiment of the present invention,
FIG. 1 is a block diagram of a connection control device, FIG. 2 is a wiring diagram showing a tree-like hierarchical structure, FIGS. 3 to 7 are schematic diagrams showing its basic operation, and FIG. 8 is a block diagram of a node device.
FIG. 9 is a block diagram of the module, and FIG. 10 is a throughput-delay time characteristic diagram. 1... Node device, 1d... Upper direction signal line, 1e
... Lower direction signal line, If ... First-come-first-served identification signal line, 2
... Transmitting/receiving terminal, 9... Failure detection department Applicant: Ricoh Co., Ltd. -, 71 quantity
Fig. 18, 3 L4- b

Claims (1)

[Claims] 1. Having a single upper direction port and a plurality of lower direction ports, the first lower direction input port is connected to the upper direction output port according to first-come-first-served logic, and the upper direction input port is connected to the lower direction direction port. A single node device with a function to connect to an output port or multiple node devices connected in a tree-like hierarchical structure via a transmission path is installed to detect the presence or absence of a signal at the input port and transmit only when there is no signal. A plurality of transmitting/receiving terminals connected to the node device are provided with a function of allowing retransmission in the event of a signal collision, and among the signals transmitted almost simultaneously from the plurality of transmitting/receiving terminals, the highest-ranking node device transmits the signals first. In a communication network control method that broadcasts only signals into the communication network and excludes other signals, the connection between node devices includes signals from a lower node device to a higher node device, and a signal from a higher node device to a higher node device. Using at least three signal lines for transmitting a signal to a lower node device and a first-arrival identification signal from an upper node device to a lower node device, an upper direction input port and a first arrival identification signal are sent to an upper direction port of a node device. It has an input port and an upper direction output port, and the lower direction port has a lower direction input port, a first-arrival identification signal output port, and a lower direction output port, and when there is a signal on any of the lower direction input ports. , the node device connects the lower direction first-come-first-served input port to the upper-level direction output port, disconnects the other lower-direction input ports, and outputs the first-come-first-served identification signal from the first-come-first-served identification signal output port; When there is a signal at the input port, the node device connects it to all lower direction output ports regardless of the presence or absence of a signal at the lower direction input port, and if there is a signal at the first arrival identification signal input port, this first arrival identification is performed. Connect the signal input port to the first-arrival identification signal output port, and when there is a signal on any of the lower direction input ports, the highest-level node device detects that lower direction first-come-first-served input port, and other lower direction inputs At the same time as disconnecting the port, connect the lower direction first-come-first-served input port to all lower-order direction output ports, and output the first-come-first-served identification signal from the first-come-first-served identification signal output port. When the identification signal detects that the signal output from the terminal does not reach the highest node device first and communication is unsuccessful, the node device to which the terminal is connected performs retransmission control after a predetermined time, or the node device to which the terminal is connected When it is detected that the signal output by itself is not the first to arrive at the highest node device due to the first-come-first-served identification signal of the first-come-first-served identification signal input port, and communication is unsuccessful, the first-come-first-served input port among the lower direction input ports is detected. A communication failure signal is output from the lower direction output port corresponding to the input port where the signal was received, and if this communication failure signal is present at the input port, the transmitting terminal determines that the communication is failure and controls retransmission after a predetermined time, The node device detects the presence or absence of a signal on the lower direction input port and the upper direction input port, detects the end of communication by detecting that the signal on both input ports disappears, and initializes the connection state of each port. 1. A communication network control method characterized in that: 2. The first-come-first-served identification signal shall indicate that the signal output by itself is a delayed input to the upper node device,
When there is a signal on any of the lower direction input ports, the node device connects the first input port to the upper direction output port, disconnects the other lower direction input ports, and connects the other lower direction input ports to the lower direction first input port. A first-come-first-served identification signal is output from the first-come-first-served identification signal output port that does not correspond, and the highest-level node device connects the upper direction output port to the upper direction input port to perform common control with other node devices, and then outputs the lower direction input. When there is a signal on any of the ports, the first input port in the lower direction is connected to the output port in the upper direction, the other lower direction input ports are disconnected, and the first arrival identification signal output port that does not correspond to the lower direction input port is connected. By outputting a first-come-first-served identification signal from
Immediately, when there is a return signal at the upper direction input port, it is output from the lower direction output port, and the node device to which the transmitting/receiving terminal is connected sends it to multiple lower direction output ports among the lower direction output ports to which the transmitting/receiving terminal is connected. If there is a signal at the input port, a communication failure signal will be output from the lower direction output port corresponding to the lower direction input port where the input signal was received, or if there is a first come identification signal at the first arrival identification signal input port, the lower direction first come first served signal will be output. 2. The communication network control method according to claim 1, wherein the communication failure signal is output from a lower direction output port corresponding to a lower direction input port where a signal including the input port has been received. 3. A communication network failure detection unit is provided in the communication network connection device of the transmitting and receiving terminal, and when this failure detection unit detects that there is no input signal at the input port within a predetermined communication network time constant starting from the start of transmission. 2. The communication network control method according to claim 1, wherein the communication network is determined to be out of order. 4. At least a failure detection section is provided in the node device to which the transmitting/receiving terminal is connected, and this failure detection section detects the signal input to the upper direction input port within a predetermined communication network time constant starting from the detection of the signal of the lower direction first-come-first-served input port. When the absence of a signal is detected, it is determined that there is a failure in the communication network, and upon this failure detection, the transmission/reception terminal connected to the lower direction output port corresponding to the lower direction first-come-first-served input port or all lower direction output ports is notified of the failure detection. 3. The communication network control method according to claim 1, wherein: 5. When the failure detection unit detects a failure, a communication failure signal is output from the lower direction output port corresponding to the lower direction first-come-first-served input port connected to the transmitting/receiving terminal or from all the lower direction output ports. 5. The communication network control method according to claim 4. 6. The node device to which the transmitting/receiving terminal is connected connects the lower direction input port to which the transmitting/receiving terminal is connected to the corresponding lower direction output port, and the upper direction input port has a signal and the first-come-first-served identification signal is present. When there is a first-arrival identification signal at an input port, all lower direction input ports to which transmitting/receiving terminals are connected are disconnected from the corresponding lower direction output ports, and the upper direction input ports are connected to this lower direction output port. If the first-arrival identification signal input port does not have a first-arrival identification signal and the upper direction input port has a signal, the corresponding lower direction output of the lower direction input port to which the transmitting/receiving terminal other than the lower direction first arrival input port is connected. 3. The communication network control method according to claim 1, wherein a return connection to the port is disconnected and an upper direction input port is connected to its lower direction output port. 7. When the failure detection unit of the node device detects a failure, it loops back to the corresponding lower direction output port of the lower direction first-come-first-served input port to which the transmitting/receiving terminal is connected or all the lower direction input ports to which the transmitting/receiving terminal is connected. 7. The communication network control method according to claim 4, wherein the communication network is disconnected. 8. Connect a communication network monitoring device with a higher priority than other transmitting/receiving terminals to the upper direction port of the highest node device,
The highest-level node device internally loop-connects its higher-level direction output port to at least the upper-level direction input port, and when the upper-level direction input port receives an input signal from the communication network monitoring device, the highest-level node device 3. The communication network control method according to claim 1, wherein the connection is disconnected and the upper direction input port is connected to all the lower direction output ports.