JPS6348933A - Loop communication system - Google Patents

Abstract

PURPOSE:To realize the transmission by a bus form network architecture communication rule such as token bus system or the like without complicating the constitution by outputting an idle signal from a node station whose data transmission is finished. CONSTITUTION:Node stations 1, 2 ... A, B ..., N-1, N or the like of the loop communication system send a data to a transmission line and when the transmission is finished, an idle signal is sent to the transmission line. Through the detection of the idle signal, the idle state of the transmission line is discriminated and the transmission by the bus type network architecture communication rules such as the token bus system is realized without complicating the constitution.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a communication system that efficiently uses transmission paths in a network that interconnects various computers, terminal devices, etc. and can mutually transmit information. It is. More specifically, the present invention relates to a loop communication system that performs transmission in a loop network according to a communication protocol for a bus network architecture.

BACKGROUND ART Known communication methods standardized by the IEEE802 committee include the LLC method, C3MA/CD method, and token bus method. A token ring method is also known. The token bus method uses a token passing method as a transmission path access method in a bus type network architecture, and the token ring method uses a token passing method as a transmission path access method in a loop type network architecture. What they have in common is that so-called tokens are circulated together.

To give an overview, in the token ring system, the network architecture is a loop type, so a signal train is transmitted in one direction along a loop transmission path, and each node regenerates and repeats the received signal train. Therefore, the tokens circulate through the nodes in the order in which they are connected to the transmission path. Then, the node that has the transmission request monitors the circulation of the token while reproducing and relaying the received signal sequence, and when it receives the token, it compares the priority of its own transmission data with the priority specified by the token, If the priority of its own transmission data is equal to or higher than the priority specified by the token, it asks for transmission rights and transmits the data. After the data transmission is completed, the token is sent to the transmission path.

On the other hand, in the token bus method, since the transmission line network architecture is in the form of a bus, it is not possible to physically circulate tokens, but instead a logical ring is created between nodes connected to the bus to sequentially transfer tokens. I'm trying to pass it on.

That is, when the bus is silent, a node station seeking transmission rights issues a MAC control command "claim-token." Then, as shown in FIG. 3, if two nodes A and B issue "claim-tokens", a conflict occurs. Therefore, these nodes wait for a predetermined response time (for example, one slot time) after completing the transmission. Each node seeking the right to transmit issues a "claim-token" again after confirming that the bus has become silent. this"
The retransmission of the "Claim Token" is repeated as the priority of the transmission request is higher. At that time, if the "Claim Token" issued by two nodes A and B conflict, the predetermined response time will be repeated again. Wait (for example, 1 slot time). Then, when there is no signal on the bus, the node seeking the right to transmit issues a "claim-token'" again and again, and if node B has stopped transmitting at that time, a conflict will occur, and node Δ acquires the right to send.

A node station that newly joins a logical ring receives a MAC control control command "follow" (who-f o 1 low) or "solicit-successor" (solicit-succ) from the token holding node.
When ``essor'' is issued, respond with ``set-successor'' (set-successor). “
If only one node responds with "Set Successor", it will give the token to that node. However, if multiple nodes issue "Set Successor", the token holding node will not respond with "Resolve Container". tension"(
resolve-contention),
Adjust the competition. Thereafter, after waiting for a predetermined response time (for example, one slot time), if only one node responds with "Set Successor", that node is given a token.

Problems to be Solved by the Invention However, the token bus system was developed for the purpose of a bus-type network architecture in which nodes are connected by coaxial cables or twisted pair lines. Therefore, when applied to optical communications, the following problems arise.

That is, in order to realize a bus-type network architecture in optical communication, a star coupler method is adopted. In that case, the token bus system can be implemented as is. However, in the case of star couplers, the number of nodes that can be connected is limited. Furthermore, the distance between the star coupler and the node is limited because the losses within the star coupler are relatively large. In addition, the star coupler itself is expensive.

Therefore, when constructing a network architecture using optical communications, a loop format must be adopted unless the number of nodes is small.

When the token bus method is applied to the loop-type optical communication network, signals transmitted once circulate through the optical transmission path. Therefore, there is a problem that a no-signal state cannot be conveniently created. For example, as shown in Figure 3, if there is a conflict between "claim-tokens," even if you wait a predetermined response time (for example, one slot time), a portion of the "claim-tokens" will continue to circulate through the optical transmission line. Therefore, the bus does not have no signal. If no signal on the bus can be detected, retransmission of the "claim-token" for conflict resolution cannot be implemented. Furthermore, even if Node B stops transmitting as a result of conflict resolution, some of the "claim tokens" that Node B previously sent are still circulating on the optical transmission line, and in this case as well, the bus becomes silent. It won't happen. Therefore, the token bus method cannot be directly applied to a loop type network.

Therefore, for example, when the number of nodes in a bus-type optical communication network using star couplers increases and it is modified to a loop-type optical communication network, or when converting an existing bus-type network using coaxial cables or twisted pair lines into an optical communication network. When remodeling to a bus-type network architecture such as a token bus method, it is not possible to use the communication method of a bus-type network architecture as is. Therefore, it is possible to change to a token ring method, but this would require a substantial change in the architecture of each node. Requires change and is not easy.

SUMMARY OF THE INVENTION Therefore, the present invention aims to provide a loop communication method that allows transmission in a loop type network according to communication rules for a bus type network architecture with slight modifications.

Means for Solving the Problems According to the present invention, in a loop-type network configured by physically connecting a plurality of 7-bands in a loop-like manner via a transmission line, it is possible to connect to the loop-shaped transmission line. A node that has finished transmitting data outputs an idle signal to the loop-shaped transmission path to notify each node that there is no data to be transmitted on the loop-shaped transmission path. Transmission is carried out by

Effect As described above, by intentionally outputting an idle signal to the loop-shaped transmission line after data transmission is completed, each node station detects the idle signal, and thereby
It is possible to know that the node station that was transmitting data has finished transmitting data. Therefore, since it can be determined that the transmission path is in an idle state by detecting the idle signal, transmission can be performed according to the communication protocol for a bus-type network architecture such as the token bus method.

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a schematic diagram of a loop-type network in which the invention may be implemented. For example, multiple node stations 1.2, . . . are connected via a transmission line such as an optical fiber.
A, B, . . . N-1, N are interconnected in a loop. The signal X sent through the transmission path is regenerated and relayed at each node station, and is transmitted clockwise, for example, as shown in FIG.

In such a loop-type network architecture, as illustrated in FIG. 2, data transmitted by a node station A is relayed at each other node station and loops back to the sending node station A. Then, for example, the receiving node station B takes in the received data into its internal buffer while relaying it.

When node station A finishes transmitting its data, it transmits an idle signal. Each other node station receiving that idle signal is node station A.
The end of transmission can be recognized.

This idle signal is transmitted, for example, during the time that the signal goes around the loop (the total time of the transmission time of the loop transmission path and the delay time at each node station). In this way, transmission data is completely eliminated from the loop transmission path. Therefore, the transmitting node does not need to check the received data while transmitting the idle signal.

After detecting the idle signal, the next node station to transmit data starts transmitting data, overwrites the idle signal with the transmitted data, and erases the idle signal.

FIG. 2 schematically illustrates the above state.

Note that the idle signal is not only transmitted after a frame in which normal information is transmitted, but also after a "claim-token" frame as shown in FIG. By transmitting an idle signal after the "claim-token" frame in this way, even if there is a conflict between nodes that wish to acquire the transmission right when there is no signal, the conflict can be resolved.

Further, the length of the minimum unit of the idle signal that enables identification of the idle signal is set to be shorter than the time required for one round of the signal loop.

Furthermore, the idle signal is created with a code different from the code used for data transmission so that the idle signal can be easily distinguished. As an extreme example, a signal that violates the code rules for digital transmission is considered an idle signal. Or 48
A 5B code is used for data transmission, and a code that is not used for data transmission is assigned to an idle signal.

In the above embodiment, the idle signal is transmitted for the necessary time for the idle signal to go around the loop transmission path, but the node transmitting the idle signal (that is, the node that has finished transmitting data) is When detecting the start of transmission, or when confirming that the end of the last transmission data transmitted by itself has gone around the loop transmission path, the transmission of the idle signal may be stopped and the relay state may be established. In either case, the node that newly started data transmission will overwrite the idle signal circulating on the loop transmission path with data and transmit the data.
This is similar to the embodiment of the present invention described above.

Note that when stopping transmission of an idle signal when detecting the start of transmission by another node, the node transmitting the idle signal must use Grf 3n to confirm that the received data is transmission data from another node. Therefore, it is necessary to check the source address in the received frame. Therefore, each node needs to be provided with means for checking the source address in the received frame.

Furthermore, when a node station that has obtained transmission rights and transmits data in packets has a large amount of data to transmit, it is customary to divide the data into a plurality of packets and transmit the data. In that case, the sending node outputs a busy signal different from the idle signal to the transmission path from the end of transmission of one packet until the start of transmission of the next packet to indicate that the transmission path is in use. is preferred.

Effects of the Invention As is clear from the above description, according to the loop communication method according to the present invention, the communication method of the communication protocol for the bus-type network architecture can be applied to the loop-type network with extremely small changes.

In particular, the present invention is extremely effective for optical communication networks where it is physically difficult to adopt the HSS format when optical fibers are used as transmission paths.

[Brief explanation of the drawing]

FIG. 1 is a diagram illustrating the basic configuration of a loop type network to which the loop communication method according to the present invention can be applied. FIG. 2 is a diagram illustrating the flow of transmission data in one embodiment of the loop communication system according to the present invention. FIG. 3 is a diagram illustrating a method for resolving conflicts between "claim-token" frames in the token bus system. [Main reference numbers] 1.2,...A, B,...N-1, N...Node station

Claims (11)

[Claims] (1) In a loop-type network configured by physically connecting multiple nodes in a loop-like manner via a transmission path, a node that has finished transmitting data to the loop-like transmission path returns to the loop-like transmission path. A loop communication method characterized in that transmission is carried out according to a communication protocol for a bus-type network architecture by outputting an idle signal to notify each node that there is no data to be transmitted on the loop-shaped transmission path. (2) The loop communication method according to claim (1), wherein the node that has finished transmitting transmits the idle signal for a necessary time for the idle signal to make one circuit around the transmission path loop. . (3) The loop communication system according to claim 1, wherein the node that has finished transmitting stops transmitting the idle signal when it detects that another node starts transmitting. (4) Claims characterized in that the node that has completed transmission stops transmitting the idle signal when it confirms that the end of the final transmission data that it transmitted has completed one circuit of the transmission path loop. The loop communication method described in paragraph (1). (5) The length of the minimum unit of the idle signal that enables discrimination of the idle signal is shorter than the time required for the signal to go around the transmission line loop. ) to (4). (6) The loop communication method according to any one of claims (1) to (5), wherein the communication protocol for bus-type network architecture is a token bus method. (7) The loop communication system according to claim (6), wherein the transmission data is a frame containing transmission information. (8) The loop communication system according to claim (6), wherein the transmission data is a claim token frame. (9) According to any one of claims (1) to (8), wherein the idle signal is created using a code different from a code used for data transmission. Loop communication method described. (10) A node that obtains the transmission right and transmits multiple packets outputs a busy signal different from the idle signal to the transmission path from the end of transmission of one packet until the start of transmission of the next packet. A loop communication system according to any one of claims (1) to (9). (11) The loop communication system according to any one of claims (1) to (10), wherein the transmission path is an optical fiber.