JPH07235924A - Monitoring method for communication network - Google Patents

Abstract

PURPOSE:To monitor the normality of a communication network by the simple hardware/software. CONSTITUTION:At a node 1, an identifier which recognizes the transfer direction of a test frame and an identifier which recognizes the node to be transferred next for the node that received the test frame are added to the information showing the connecting relation of the link to be tested, and the test frame is constructed, and sent to a node 2. At the node 2, an identifier which recognizes a node to transfer the received test frame is changed so that the node that received the test frame can recognize the node to be transferred next, then the test frame is transferred to a node 3. The node 2 writes a time stamp in the received test frame and changes the identifier which recognizes the transfer direction of the test frame to return it as a response frame to the node 1 where the test frame is transferred. The node 1 receives the response frame from the node 2 and checks the time stamp by an inspecting function. Thus the transfer performance is checked between both nodes 1 and 2.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication network monitoring method, and more particularly to a communication network monitoring method for performing label multiplex communication.

Digital multiplexing methods include synchronous multiplexing and asynchronous multiplexing. Synchronous multiplexing has been the mainstream technology since the development of digital communication, and it is based on the relative position information of the frame start position with a constant cycle and the time slot whose phase is always synchronized. Is a method of performing. That is, it can be said that it is a position multiplexing system. On the other hand, the asynchronous multiplex system is a system in which the relative position to the frame phase of the network is not fixed, and the multiplex / separation is performed by recognizing the label added to the data.
That is, it can be said that it is a label multiplex system.

Communication using the label multiplex system includes packet communication and ATM (Asynchronous Transfer Mode) communication. The packet communication system has been put into practical use for a long time, and is progressing to a frame relay in which the protocol is simplified. This frame relay is just beginning to spread in Japan. Further, the ATM communication technology is a basic technology of B-ISDN (broadband ISDN) called a next-generation communication network, and will be in full swing technically and in service. And in the future, frame relay and A
It is considered to be a fusion with TM. The present invention simplifies the monitoring method in a communication network that performs label multiplex communication in such a status. In the following, the frame relay will be described as an example.

In the frame relay technique, a transmitting terminal forms a frame in which a flag formed by a specific code pattern is attached before and after a field consisting of control data and information, and sends the frame to the network. The network transfers the frame and delivers it to the receiving terminal while referring to the flag and the control data. The receiving terminal disassembles the frame and extracts the communication information.

In this way, for example, high-speed data transmission between computers is performed, but at present, an existing digital network is used as a physical line rather than a network dedicated to frame relay being established. Therefore, in addition to the function unique to the frame relay, the frame relay network needs to have a function of multiplexing and transmitting a frame in the existing network and separating the frame from the existing network. That is, existing networks simply provide a physical line for sending frames,
It is transparent to the frame.

[0006]

2. Description of the Related Art An existing network has a function of monitoring the normality of the network itself, but is not transparent to a frame, that is, it does not check or process the content of the frame. So
Even if it is confirmed that the existing network is normal, it is not possible to confirm whether or not the frame is transmitted / received correctly.

Further, whether or not a node of the frame relay network is equipped with a standard management protocol is considered to be an exclusive matter of the user. Therefore, the frame relay network itself is not always monitored by the standard protocol. Absent.

In such an environment, when the frame relay network is started up or the maintenance of the frame relay network is performed, it may be necessary to confirm the normality of the frame relay network by another confirmation means. Since the start-up and maintenance of the frame relay network are usually the business of the manufacturer or dealer, the confirmation means necessary therefor does not impose a financial burden on the user.

Therefore, the means for confirming the normality of the frame relay network must be light in terms of both hardware and software.

[0010]

SUMMARY OF THE INVENTION The present invention addresses such a problem and provides simple hardware for a communication network for label multiplex communication such as a frame relay network.
It is an object to provide a method of monitoring the normality of communication that can be realized by software.

[0011]

FIG. 1 is a principle diagram of a test according to the present invention. In FIG. 1, 1 to 5 are nodes. Each node has a main processor 11, 21, 3
1, 41, 51 are provided, and each main processor has a test frame transfer function and a response frame verification function that is a response to the test frame. Here, in FIG. 1, the node 1, the node 1, the node 2,
Since the case where the link having the node 3 and the node 4 is tested is illustrated, only the node 1 is shown as having a verification function.

[0012]

In the node 1, the information indicating the connection relationship of the link to be tested, the identifier for recognizing the direction in which the test frame is transferred, and the node (node 2 in this case) that received the test frame should transfer next. An identifier for recognizing the node is added to form a test frame, and the test frame is transferred to the node 2. The node 2 changes the identifier for recognizing the node to which the received test frame is to be transferred so that the node receiving the test frame (in this case, the node 3) can recognize the node to be transferred next, Transfer the test frame to node 3. Also, the node 2 writes a time stamp in the received test frame, changes the identifier for recognizing the transfer direction of the test frame, and returns it as a response frame to the node that has transferred the test frame (node 1 in this case). To do. The same operation is performed in the nodes 3 and later.

Upon receiving the response frame returned from the node 2, the node 1 checks the time stamp by the verification function to check the transfer performance of the link between the nodes 1 and 2. Also, when the response frame is not returned from the node 2 even after the lapse of a predetermined time,
The verification function of node 1 also checks this.

Although the above is a partial explanation, since all the nodes operate in the same manner, the verification function of the node 1 can monitor the link communication and the link transfer performance.

[0015]

FIG. 2 is an image diagram of a node. In FIG. 2, 11 to 14 are communication processors, 15 is a main processor, 16 is a storage device, 17 and 18 are buses, and 19
Is the console.

Here, the main processor is provided with a transfer function for transferring the frame used for the test and a verification function for verifying the returned test frame. It should be noted that FIG. 2 shows only the frame relay communication function, and the function of multiplexing the frames and sending them to the transmission path, separating the signal received from the transmission path and decomposing it into frames is omitted. .

The communication processor detects a frame input from the input path and writes the detected frame in the storage device via the bus. The main processor refers to the address added to the frame written in the storage device, passes it to the communication processor of the outgoing route, and sends the frame to the outgoing route by the communication processor. Although an accurate frame structure is not shown in FIG. 2, the communication processor 11 receives a frame having an address X, sends the frame from the communication processor 14, and the communication processor 13 receives a frame having an address Y. An example is shown in which the communication processor 12 sends out the frame.

In the node, the processor to be transferred acquires the bus and transfers the frame to an appropriate transfer destination. The console also sends and receives frames with basically the same structure as the frames that request the bus for communication,
The frame is used for testing.

The frame output from the console specifies an identifier indicating a test frame, an identifier indicating the transfer direction of the test frame, and an ingress route and an egress route of a node on a communication channel to be tested. Have a table and
The main processor activates the transfer function and transfers the test frame to an appropriate transfer destination by referring to the transfer direction identifier and the table that specifies the ingress path and the egress path of the node on the communication path to be tested. . The node that receives this test frame similarly transfers it to the next node, changes the transfer direction identifier, and returns it as a response frame. When the node which has transmitted the test frame receives the response frame, the main processor activates the verification function to determine whether the test frame is normally returned.

FIG. 3 shows the structure of a test frame. The horizontal direction of the drawing is 8 bits, which is the number of bits forming an octet.
The vertical direction is the number of octets. The test frame has a 1-octet flag at the beginning and a 1-octet flag at the beginning, and a data link connection identifier (DLCI) consisting of 2 octets after the flag at the beginning. It has a Frame Check Sequence (FCS) of 2 octets. DLCI and F
An information field is provided between CSs, and the information field is composed of n combinations of a frame direction indication octet and a route specification table pointer octet, and octets of an ingress and egress route number and a plurality of octets for writing a time stamp. .

FIG. 4 is a processing flow of the test frame. The operation will be described below in the order of reference numerals. A. The node receives the frame.

B. The value that DLCI means test (Fig. 4
Then, it is confirmed whether it is "0000000001"). C. When DLCI is "0000000001" (Ye
In s), the input / output route number stored in the octet of the route designation table having the same number as the pointer value indicating which octet in the route designation table should be read is read.

D. It is determined whether the read out route number is the end point code. When it is the end point code (Yes), there is no node to be transferred before, so the process jumps to G.

E. If it is not the end point code (No), the route designation table pointer is incremented by +1 and the time stamp is written. F. The test frame in which the designated table pointer is changed and the time stamp is written is sent to the route having the route number read in C.

G. It is determined whether or not the incoming route number read in C is the starting point code. When it is the starting point code (Ye
Since there is no node for sending the response frame in s), the process ends.

H. If it is not the start point code, the route designation table pointer is decremented by 1, the time stamp is written, and the frame direction instruction is changed to a value indicating the response frame.

I. The response frame is sent to the designated entry route and the process ends. J. If the DLCI is not “0000000001” in B, normal frame relay processing is performed.

It will be understood from the above description how the transfer function works, but a concrete description will be given with reference to FIGS. 5 and 6. Before that, define the pointers used in the information fields.

The frame direction indication is "0000000".
"0" means a test frame, "00000001"
Means a response frame. The route number is arbitrarily determined for each node, and the number is assumed to be the number displayed beside the link entering and leaving the node in FIG. For example, the link between the node 1 and the node 2 is defined as "3" in the node 1 and "2" in the node 2. Moreover, this route number is defined to be the same "3" for the node 1 to the node 2, for the outgoing link seen from the node 1, entering from the node 2 to the node 1 and for the incoming link seen from the node 1. . The route number thus defined is written in the octet storing the route number. However, the starting point is "0000000
0 "and the end point are displayed as" 11111111 "(" FF ").

FIG. 5 is a diagram (part 1) for explaining the delivery of the test frame in the node 1. In FIG. 5, the area for writing the time stamp is omitted. FIG. 5A shows a test frame generated by the node 1. The flag is "7E", which is internationally standardized. Therefore, when represented by "0" and "1", "0111111"
The DLCI uses 10 bits out of 16 bits, and here, "0000000001" is used to indicate that it is a test frame. The frame direction indication is defined as a test frame. As described above, it is set to “00000000.” The route designation table pointer is
First, set to "00000000". As a result, the "0" octet in the route designation table is read, and the outgoing route number "0" is read.
011 ”and the incoming route number“ 0000 ”are read in. Since the incoming route number is the start point code, no response frame is generated.

When the generated test frame is transferred to the node 2, the route designation table pointer is incremented by 1 to change to "00000001" as shown in FIG. To do.

FIG. 6 is a diagram (part 2) for explaining the delivery of the test frame in the node 2. Also in FIG. 6, the area for writing the time stamp is omitted. Of course, the input frame in FIG. 6 (a) has exactly the same contents as in FIG. 5 (b). Since the route designation table pointer of the received frame is "00000001", the "1" octet of the route designation table is read and the output route is "0100".
At ("4"), it is recognized that the incoming route is "0010"("2"). Therefore, as shown in FIG. 6B, “1” is added to the route designation table pointer of the received test frame to make it “00000010” (“2”), and the time stamp is written and sent to the node 3. To do.

On the other hand, the node 2 sends a response frame to the node 1. At this time, as shown in FIG.
"000" indicating the direction of the frame indicating the response frame
It is rewritten to "00001", and the pointer value "00000001"("1") at the time of receiving the route designation table pointer is decremented by "-1" to "00000000"("0") and the incoming route designation "0010" is set. Refer to it and send it to the route “2”.

In this way, the test frame is transferred to the subsequent node and the response frame is returned to the previous node. At the final node (node 4 in FIG. 1), when the route number is read, the end point code is stored in the outgoing route number, so only the response frame is returned.

The node 1 checks the number of response frames and the time stamps returned one after another to verify the performance and normality of the communication network. FIG. 7 is a flowchart of the verification function. The operation will be described below according to the order of the symbols. L. A counter that counts the number of response frame receptions is reset to zero. M. Receive the frame. N. It is determined whether the DLCI is "0000000001". If it is not "0000000001" (No), normal frame relay processing is performed (U). P. When the DLCI is “0000000001” in N (Yes), since the response frame is received, the counter is incremented and the time stamp is stored. Q. It is determined whether or not a predetermined time determined by the round trip time with the far away node has elapsed. If the predetermined time has not elapsed (No), the process jumps to M and performs a stampy so that the frame can be received. R. When the predetermined time has passed (Yes), the received time stamp is analyzed and the number of times the time stamp is received is confirmed. The time stamp analysis is performed by comparing with the contents of the storage device that stores the reference time stamp to be written by each node, since it is possible to determine which node has written by the written octet number. That is, the congestion state of the link is determined by whether the time stamp is earlier or later than the reference time stamp stored in the storage device. If the number of receptions is less than the number of test target nodes, it can be immediately determined that the communication of the network is not normal, and the fault position can be estimated by the octet in which the time stamp is not written. For example, in FIG. 1, if the response frame in which the region where the node 2 should write the time stamp is filled is received and the response frame where the region where the time stamp is written is filled in by the node 3 and node 4 is not received, It can be seen that there is a failure after the link between nodes 2 and 3.

In this case, the failure position cannot be specified, but in this case, the same test can be carried out with the node 4 as the starting point and the node 2 as the ending point for further details. That is, at this time, if the node 3 returns the response frame and the node 2 does not return the response frame at this time, considering that the node 2 returns the response frame when tested from the node 1, It can be specified that there is an abnormality in the link between 2 and node 3.

As described above, the normality of the frame relay network can be tested. Although a time stamp is written as an identification signal to be written when a test frame or a response frame is transmitted here, an identification signal such as a node number may be used instead of the time stamp.
However, if a time stamp is written, the time from when a test frame arrives at a node to when a response frame is output and the transfer time between nodes are known, so it is possible to directly check the performance and status of the communication network. Is advantageous.

The frame relay network has been described above as an example. However, even in other communication networks using label multiplexing, even if the label structure is different, there are areas corresponding to the flags and DLCI described in FIG. , A similar test function can be realized.

[0039]

As described above, in a communication network that performs label multiplexing, it is possible to easily test communication of the communication network without preparing complicated hardware and software.

[Brief description of drawings]

FIG. 1 Principle diagram of the test.

FIG. 2 is an image diagram of a node configuration.

FIG. 3 is a structure of a test frame.

FIG. 4 is a processing flow of a test frame.

FIG. 5 is a diagram (part 1) explaining delivery of a test frame.

FIG. 6 is a diagram (part 2) explaining delivery of a test frame.

FIG. 7 is a flowchart of the verification function.

[Explanation of symbols]

 1, 2, 3, 4, 5 nodes 11, 21, 31, 41, 51 Main processor Test frame transfer function Response frame verification function

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical indication location H04L 29/14 H04Q 3/00 9077-5K H04L 11/20 102 Z 9371-5K 13/00 315 Z

Claims (4)

[Claims] 1. A method for monitoring a communication network in a communication network for performing communication by label multiplexing, wherein a signal for testing the communication network is created and transferred to a node forming the communication network,
The signal received from upstream is first processed and transferred further downstream, the received signal is processed second and transferred upstream, and the signal transferred from downstream is processed third. It is equipped with a transfer function that transfers the data upstream and a verification function that analyzes the signal transferred from the downstream to determine the normality of communication in the communication network. Create a signal, transfer it downstream, analyze the signal transferred from the downstream node, and perform the first processing on the signal received from upstream at the node under test at the node under test. The signal received from the upstream, and then the second processing to the upstream, and the signal transferred from the downstream to the third processing to the upstream, At the target node, the end of signal transfer Position in the node, a monitoring method of a communication network, characterized in that the signal received from the upstream to the second processing transfers to upstream. 2. The method for monitoring a communication network according to claim 1, wherein the signal for testing the communication network has a flag at the beginning and at the end of the signal, and between the flag at the beginning and the flag at the end. At least, the display data of the test function, the data indicating the direction in which the signal is transferred, the route specification table defining the connection of the network to be tested, and the signal processed by the node receiving the signal are transferred. Monitoring of a communication network, which is a signal having a pointer for reading a power node from a route specification table and an area for writing an identification signal when a signal processed by a node receiving the signal is transferred Method. 3. The method of monitoring a communication network according to claim 1, wherein the first processing is a pointer for reading from a route designation table a node to which the node that received the signal should transfer the processed signal. The second process is to change the data indicating the signal transfer direction, and to read from the route specification table the node to which the node that received the signal should transfer the processed signal. Is a process for changing the pointer in the direction opposite to the first process, and the third process is for reading from the route specification table the node to which the node that received the signal should transfer the processed signal. A method for monitoring a communication network, characterized in that the pointer is changed in the opposite direction to the first processing. 4. The communication network monitoring method according to claim 1, wherein the verification function stores at least an identification signal written in a signal transferred from a downstream side in a storage device. A method for monitoring a communication network, which comprises comparing with an identification signal to be transferred from a downstream side.