JPH0683260B2 - Detour method - Google Patents

Description

The present invention relates to a detour method of a packet switching system for exchanging packets between originating and terminating processes of arbitrary originating and terminating nodes, and promptly detects the presence or absence of an invalid pending resource due to a poor quality state of a transmission line. For the purpose of effective use of resources in the packet switching system, the programs of any originating and terminating nodes are connected, the outgoing route is selected for each packet based on the node information, and the resources are used for each node. In the detour method of the packet switching system provided with the quantity monitoring unit, the transmission path failure monitoring unit and the routing mechanism, when no response occurs n times or more within a predetermined Ta ₁ hour, or the packet is sent to the transmission queue corresponding to the line. When a predetermined Tb ₂ hours have elapsed after the connection, the bypass request is transmitted when the connection is suspended, and the predetermined Ta ₂ is transmitted after the bypass request is transmitted.
A transmission path quality monitoring means is provided for sending a detour cancellation request to the routing mechanism after a lapse of time or after a lapse of a predetermined Tb ₃ hours, and when the detour request is sent to the routing mechanism, the routing table is moved from the optimum route. It is configured to switch to the detour route and return the routing table to the optimal route when the detour cancellation request is sent.

[Industrial application field]

The present invention relates to a detour method of a packet switching system for exchanging packets between originating and terminating processes of an arbitrary originating and terminating node.

A packet switching system stores a message from a terminal in a switch, divides it into short data blocks of, for example, about 1000 bits, edits the packet with headers such as a destination and a serial number, selects an outgoing route, and selects the next switch. , Or to the terminal, but at this time, it is necessary to effectively use the resources of the packet switching system.

[Conventional technology]

FIG. 6 shows a configuration diagram of a packet switching system for implementing the conventional example.

In the figure, nodes A, B, C which are mutually connectable via transmission lines a, b, c and terminals X, which are capable of mutually communicating with nodes A, B, C via transmission lines x, y, z. Y, Z are shown.

A physical layer control mechanism 20, which activates, maintains and deactivates a physical connection for transmitting bits to each node and provides mechanical, electrical, functional and procedural means,
A data link layer control mechanism 21 that provides a functional and procedural means for establishing, maintaining, releasing a data link, and transferring data, establishing a connection between processes Pa to Pc that are parties involved in packet communication, A packet processing mechanism 22 that maintains and releases, scrutinizes individual packets, checks the status, and performs transmission / reception processing, a routing mechanism 23 that manages the state of the line and selects the optimum route for packet transmission, and fault monitoring and transmission of transmission paths. From the detour condition detection mechanism 24, etc. that monitors resource usage, such as ensuring that the queue length of packets to be sent for each route does not exceed a certain value, and monitors detour condition occurrence during routing. It is configured.

The linkage between processes includes a virtual call (VC) performed in the process of setting up a call and a permanent virtual circuit (PVC) fixedly set in advance.
Here, since the configuration diagram of each node is the same, only node B is shown.

Further, the terminal may be a packet terminal or a non-packet terminal. In the latter case, the packet assembly & disassembly (PA) is installed on the node side.
D) is required.

FIG. 7 is an explanatory view of the routing processing process of FIG. The routing process will be described below with reference to FIG. 7 with reference to FIG.

While the physical connection and the data link are established between the physical layer control mechanism 20 and the data link layer control mechanism 21 in the node B and the terminal Y, the packet 1 addressed to the terminal Z is first transmitted from the terminal Y. When transmitted to the node B via the path y, the data link layer control mechanism 21 causes the frame synchronization flag F, the address field (A), the control field (C), the information field (I) and the frame transmission error detection frame. An I (information) frame on a high level data link control procedure (HDLC) including a check sequence (FSC) and the synchronization flag (F) is received.

The packet 1 is included in the information field of this I frame.

Further, the packet processing mechanism 22 checks the packet, that is, checks the format, and at the same time checks the state of the process Pb corresponding to the packet to determine whether the packet can be received. Mechanism 23
Selects the transmission path b, which is the optimum path for the destination node C, from the routing table.

In this case, assuming that the detour condition has not occurred for the transmission line b, the packet 1 is transmitted to the node C via the transmission line b and further transmitted to the terminal Z via the transmission line z (see FIG. 7-
reference).

After that, when a physical layer abnormality occurs on the transmission path b, and the transmission path failure monitoring unit 241 (collecting all the information about the line failure) in the detour condition detection mechanism detects this, the routing mechanism
Since the detour condition occurrence (line failure occurrence) is notified to 23, the routing mechanism of the node B turns on the line failure occurrence display corresponding to the transmission path b in the routing table, and the transmission path a which is the detour is selected. Like

Therefore, when the packet 2 arrives from the terminal Y, the routing mechanism selects the transmission path a, which is a detour, and the packet 2 is transmitted to the node A via the transmission path a.

In the node A that has received the packet 2, the routing mechanism 13 confirms that the destination is the node C by looking at the inter-office header added to the packet 2, and relays this packet to the node C via the transmission path c. Therefore, from node C to transmission line z
The packet 2 is sent to the terminal Z via the (see FIG. 7-).

Next, when the transmission path failure monitoring unit 241 of the node B detects the failure recovery of the transmission path b, it notifies the routing mechanism 23 of this, so the routing mechanism 23 returns the routing table to the original state, and the packet 3 is the same as the packet 1. It is transferred by a route (see Fig. 7-).

After that, with respect to the transmission path b, the resource use amount monitoring unit 242 in the detour condition detection mechanism of the node B is checked for resource use amount excess such as packet queue length excess and throughput excess.
When it is detected, the routing mechanism 23 is notified of the occurrence of the bypass condition (resource usage excess).

Therefore, the routing mechanism turns on the resource usage excess occurrence display for the transmission path b in the routing table,
Since the transmission path a which is a detour is selected in the same manner as in the above case, the packet 4 arriving at the node B is transferred through the same detour as the packet 2.

However, when the resource usage amount returns to the normal value, the routing table is returned to the original state, so that the packet 5 arriving thereafter is transferred along the normal route in the same manner as the packets 1 and 3 (see FIG. 7-).

[Problems to be solved by the invention]

However, an RR (receive ready) frame indicating that the transmitted I frame was correctly received by the node C due to the occurrence of an FSC error caused by a momentary interruption or noise on the transmission path b Cannot be received even after the lapse of a predetermined time T ₁ .

That is, when packets 6 and 7 arrive at the node B in a state in which the protocol related to the data link layer frequently occurs such as occurrence of response waiting timeout and the quality of the transmission line is deteriorated, these packets are Since normal transfer cannot be performed via the transmission line b and the detour is not selected, the node B may stay for a long time.

This is not detected as abnormal by the transmission line failure monitoring unit 241 because the abnormal state of the transmission line b is not caused by a long time interruption but by a momentary interruption or noise, and the queue length over, throughput, etc. Since the resource usage is not exceeded, the resource usage monitoring unit 242 does not detect any abnormality. However, the data link layer control mechanism detects the response wait time-out and repeats the protocol retransmission process to recover from this.

Normally, the abnormality detected by the data link layer control mechanism 21 is notified in the form of an alarm message to the operator, but unlike the case of a line failure, it may be remedied by retransmission at the protocol level. It is difficult to determine whether or not the line should be blocked and the loopback test should be performed, and whether or not feedback should be applied to the routing mechanism 23.

That is, since the detour condition detection mechanism 24 does not have a transmission line quality monitoring mechanism, when the transmission line quality deteriorates, packet transfer delay increases, resources in the packet switching system are suspended, and this suspension is canceled. In order to do so, there is a problem that it is necessary to rely on the judgment of an advanced operator.

[Means for solving problems]

The above problems can be solved by the detour method of the packet switching system shown in FIG.

Packet switching that connects processes of arbitrary originating and terminating nodes, selects an outgoing route for each packet based on node information, and provides each node with a resource usage monitoring unit, transmission line fault monitoring unit, and routing mechanism In the system detour method, 36 is invalid when no response occurs n times or more within a predetermined Ta ₁ hour, or even when a predetermined Tb ₂ hours elapse after the packet is connected to the transmission queue corresponding to the line. When it is suspended, it is considered that the quality of the transmission line is temporarily degraded, and a bypass request due to no response is generated or a bypass request due to invalid suspension is transmitted, and a bypass request due to no response is transmitted. _two hours after a predetermined Ta from, or the invalid holding avoidance request delivery to a predetermined Tb ₃ hours after the due, detour request the routine bypass cancellation request in the transmission line quality monitoring means for sending to said routing mechanism When sent to the grayed mechanism is changed to detour path the routing table from the optimum route, when the detour cancellation request has been sent is the manner returning the routing table to the optimal path.

[Action]

The present invention monitors resource retention time in the detour condition detection mechanism,
A transmission line quality monitoring mechanism 36 including a no response occurrence frequency monitoring and a timer monitoring is provided. FIG. 2 is an explanatory view of the principle of the transmission line quality monitoring means in FIG. 1, FIG. 2 (a) is an explanatory view of the number of times of no response occurrence, and FIG. 2 (b) is an explanatory view of the principle of resource retention time monitoring. The principle of the present invention will be described below with reference to FIG.

(1) Monitoring the frequency of occurrence of no response (see FIG. 2 (a)).

The non-response occurrence frequency monitoring is started via the data link layer control mechanism when the response waiting timeout, that is, the T ₁ timeout occurrence is detected by the timer monitoring, and updates the non-response counting counter (not shown). If the contents are checked and no response occurs n times or more within a predetermined Ta ₁ hour as shown at the left end of FIG. 2 (a), it is considered that the quality of the transmission line is temporarily deteriorated. , Routing mechanism 33
After making a detour request (transmission path quality deterioration detection notification) to the data link layer control mechanism 31, the non-response count counter is initialized.

After making the above-mentioned detour request, as shown in the central part of FIG. 2 (a), after making a predetermined Ta ₂ (Ta ₂ > Ta ₁ ) time, the detour cancellation request of the transmission line is unconditionally made. The transmission line is made available again.

In the case of the right end of FIG. 2A, the detour request is not sent because the number of times of no response occurrence within a predetermined Ta ₁ time is less than n.

(2) Resource retention time monitoring (see FIG. 2 (b)).

Next, in the resource retention time monitoring, the above timer monitoring mechanism is activated at regular time intervals (Tb1) as shown in FIG. 2 (b), but the packet is not connected to the transmission queue corresponding to the line. The dwell time is checked, and if it is not normally transmitted or released within a predetermined Tb ₂ hours, it is activated and it is considered that the quality of the transmission line is temporarily degraded. A detour request (transmission path quality deterioration notification) is issued to the control mechanism 31.

Here, the transmission waiting queue is a link-level transmission waiting queue that waits for a packet to be transmitted because it cannot be sent while the line is in use, or a link level that waits for delivery confirmation at the link level after being sent to the packet line. Queue for confirmation, etc.

After avoidance request performs detour cancel request of the transmission path unconditionally after a predetermined Tb ₃ hours, again to allow the state using the transmission path.

That is, by using such a transmission path quality monitoring means, it is possible to prevent an increase in the packet transfer delay time and the generation of invalid hold resources due to the low quality state of the transmission path. Further, when a detour request is made to the routing mechanism, an alarm transmission (transmission path quality deterioration) is also performed at the same time, so that it is possible for the operator to judge whether or not the transmission path should be maintained. Become.

〔Example〕

FIG. 3 shows an example of a packet switching system configuration diagram for carrying out the present invention, FIG. 4 shows a routing processing process explanatory diagram of FIG. 3, and FIG. 5 shows an operation explanatory diagram of FIG. Where the fifth
As the detour condition detection mechanism in the figure, only the transmission path quality monitoring mechanism 36 added in the present invention will be described, and it is assumed that the same Roman numerals are connected in the flow of processing. Throughout the drawings, the same reference numerals denote the same objects, * indicates the non-response counting counter Ca
The initial settings (not shown) are shown.

Further, a non-response occurrence frequency monitoring part 361, a resource retention time monitoring part 362, and a timer monitoring part 363 in FIG. 3 are components of the transmission path quality monitoring mechanism 36.

First, the function of each functional block of FIG. 3 and the mutual relation will be described with reference to FIG.

When the data link layer control mechanism of FIG. 5 (a) is notified of frame reception via the physical layer control mechanism <event analysis>
As a result, the type of event is found to be frame reception, and I frame reception is found by checking the contents of the control field (C) in the above HDLC frame format in <frame analysis> on the left branch.

Therefore, the <format check> checks whether or not the frame length, the transmission sequence number, the reception sequence number, etc. of the I frame are correct. If OK, the <status check> determines whether the I frame can be received. Check.

If all the above checks are OK, an event of I frame reception is added, the contents of the I field, that is, the packet is passed to the packet processing mechanism of FIG. 5B, and this is activated "I".

In the packet processing mechanism of FIG. 5B, <event analysis>
As a result of event analysis in, it can be seen that I frame is received.

Then, <Packet analysis> analyzes the packet type such as call request packet or data packet, <Format check> checks the format corresponding to the packet type, and <Status check> receives the packet. If all are OK, the packet is passed to the routing mechanism 33 of FIG. 5 (e) to activate this mechanism "II".

In the routing mechanism of FIG. 5 (e), the line number for sending the packet by the normal route (optimum route) is set as the takeover information to the packet processing mechanism based on the node information.
Then, the detour conditions such as <resource usage check>, <failure occurrence check>, <quality deterioration check> are checked, and if the detour condition is not generated, the normal route is determined.

However, when the detour condition is generated, the line number of the detour route is set as the takeover information to the packet processing mechanism, and a packet transmission event is added to request the packet transmission to the line, and the packet transmission event is given in FIG. "III" for activating the packet processing mechanism of (b).

The packet processing mechanism of FIG. 5 (b) enters the central branch when packet transmission is found by <event analysis>, analyzes packet type by <packet analysis>, and sends the packet by <status check>. "IV" which checks whether or not, adds an event of I frame transmission after starting packet editing such as adding an inter-office header, and activates the data link layer control mechanism.

Therefore, the data link layer of FIG. 5 (a) edits an I frame that gives a transmission serial number, a reception serial number, etc. at the “IV” branch, and checks if the I frame can be sent to the line by <Status check>. Is checked, but if transmission is possible, I
Set the no-response monitoring timer T ₁ after sending the frame to the line. As a result, the timer monitoring shown in FIG. 5 (d) is "V" for monitoring whether or not a response is received from the partner node.

However, when transmission is not possible, the I frame is linked to the transmission queue and the Tb ₂ monitoring target display of the packet is turned on. Based on this, the timer monitoring in Fig. 5 (d) is "XII" that monitors the resource retention time.

The timer monitoring of FIG. 5 (d) monitors the T ₁ timer by the branch of “V”, and when this timer <timeout>, notifies the data link layer control mechanism of FIG. 5 (a) of the T ₁ timeout and sends the data. "VI" that activates the link layer control mechanism.

Further, in order to monitor the number of occurrences of no response, the Ta ₁ timer is monitored at the branch of “V ′”, and when <timeout> occurs, the no response counter (not shown) corresponding to the line is cleared. In addition,
Branch "V '" is a timer monitoring program is activated from the operating system (OS) at a constant period Ta _1.

When the data link layer control mechanism in Fig. 5 (a) finds that T ₁ time out in <event analysis>, it enters the central branch, activates transmission line quality monitoring, and notifies the occurrence of T ₁ time out.
I ”. Also, the retry counter is updated, and it is checked whether or not <retry out>. If it is a retry out, an alarm is sent out, and if it is not a retry out, retry processing is performed.

In the transmission line quality monitoring of FIG. 5 (c), if an event is analyzed as T ₁ timeout occurrence by <event analysis>, the counter value of the non-response count counter Ca (not shown) corresponding to the line is branched at the “V” branch. Update. At this time, if the counter value of the no-response counting counter overflows, the routing mechanism and the data link control mechanism are activated to notify the detection of the deterioration of the transmission path quality "IX", and then "VIII" to set the Ta ₂ timer.
However, if it does not overflow, it does nothing.

The timer monitoring shown in FIG. 5 (d) checks the presence or absence of the Tb ₂ monitoring target packet in order to monitor the staying resource at the Tb ₁ cycle at the “XII” branch. If there is a target packet, the cumulative time since the packet was connected to the transmission queue is updated, and when the time exceeds Tb ₂ , transmission path quality monitoring is activated and Tb ₂ timeout is notified. ". If there is no Tb ₂ monitoring target packet, or if there is a target packet and there is no Tb ₂ timeout, do nothing.

The transmission line quality monitoring in Fig. 5 (c) is Tb in <event analysis>.
_{2 When it} is known that a time-out has occurred, the routing mechanism and data link layer control mechanism are activated and "IX, XIV" is sent to notify the deterioration of transmission path quality. Then, set the Tb ₃ timer to "VII.
I ”.

When the routing mechanism of FIG. 5 (e) finds that the transmission line quality deterioration has been detected by <event analysis>, it turns on the quality deterioration display of the line and displays the line unusable. In addition, the Ta ₂ timer set or the Tb ₃ timer set is used, as shown in FIG.
When the Ta ₂ timer or Tb ₃ timer is monitored in the “VIII” branch of the timer monitoring of < ₃ > and <timeout> is reached, the transmission path quality monitoring is activated by Ta ₂ timeout or Tb ₃ timeout [X].

For the transmission line quality monitoring of FIG.
_{When the} occurrence of ₂ time-out or Tb ₃ time-out occurs, the routing mechanism is activated in the branch on the right side to notify the recovery of the deterioration of the transmission path quality "XI".

When the routing mechanism of FIG. 5 (e) detects the recovery of the deterioration of the transmission line quality in <event analysis>, it turns off the display of the deterioration of the quality of the line.

Next, the routing process of FIG. 3 will be described with reference to FIG.

In FIG. 4, the mechanism in which the packet addressed to the terminal Z from the terminal Y is transferred to the transmission path b or the transmission path a at the node B depending on the occurrence of the failure of the transmission path b or the excess of the resource usage is the same as in the case of FIG. It is the same.

When the packet 4 addressed to the terminal Z from the terminal Y arrives at the node B via the transmission path y, the routing mechanism 33 of the node B attempts to send this packet to the transmission path b which is the normal route.

However, in the non-response occurrence frequency monitoring part 361 or the resource retention time monitoring part 362, a temporary quality defect of the transmission path b (no response occurs n times or more within a predetermined Ta ₁ hour, or a packet is predetermined on the transmission queue). Of Tb ₂ hours or more) is detected, the transmission line quality monitoring means 36 notifies the routing mechanism 33 and the data link layer control mechanism 31 of the detection of the deterioration of the transmission line quality.

The routing mechanism 33 changes the routing table when notified of the detection of the deterioration of the transmission path quality. As a result, the transmission line a serving as a detour is selected, and the packet 4 and the subsequent packet 5 are also detoured to the transmission line a. Further, when the data link layer control mechanism 31 is notified of the detection of the deterioration of the transmission path quality, it collects the untransmittable packet (packet 4) and activates the packet processing mechanism 32.

When the packet processing mechanism is notified of the deterioration of the transmission path quality, the packet processing mechanism activates the routing mechanism to perform packet transmission on the detour.

The transmission line quality monitoring means 36 detects the low quality state of the transmission line b,
When the predetermined Ta ₂ or Tb ₃ hours have passed, the routing mechanism 3
Unconditionally request detour cancellation to 3 (notification of recovery of transmission path quality deterioration). As a result, the routing mechanism 33 returns the routing table to the original state and sends the next received packet 6 to the transmission path b which is the normal route.

That is, according to the present invention, the low quality state of the transmission line can be recognized at an early stage, so that the resources of the packet switching system including the bypass and the packet buffer can be effectively used. Further, the operator can maintain the transmission line by the alarm information transmitted from the detour condition detecting mechanism including the transmission line quality monitoring means 36, the transmission line failure monitoring unit 34, and the resource usage amount monitoring unit 35. Easy to maintain.

〔The invention's effect〕

As described above in detail, according to the present invention, it is possible to effectively use the resources in the packet switching system and to easily maintain the transmission path.

[Brief description of drawings]

FIG. 1 is a block diagram of the principle for implementing the present invention, FIG. 2 is an explanatory diagram of the principle of the transmission line quality monitoring means in FIG. 1, and FIG. 3 is an example of a packet switching system configuration diagram for implementing the present invention. 4, FIG. 4 is an explanatory diagram of a routing processing process of FIG. 3, FIG. 5 is an explanatory diagram of operation of FIG. 3, FIG. 6 is a packet switching system configuration diagram for implementing a conventional example, and FIG. The routing processing process explanatory drawing of the figure is shown. In the figure, 33 is a routing mechanism, 34 is a transmission line failure monitoring unit, 35 is a resource usage amount monitoring unit, 36 is a transmission line quality monitoring means, 361 is a no-response occurrence frequency monitoring part, 362 is a resource retention time monitoring part, 363 Indicates the timer monitoring part.

Claims (1)

[Claims] 1. A process for connecting a process of an arbitrary destination node,
An output route is selected for each packet based on the node information, and a resource usage monitoring unit (35), a transmission line failure monitoring unit (34), and a routing mechanism (33) are provided in each node.
When the detour condition occurrence information from the resource usage monitoring unit or the transmission path failure monitoring unit is sent to the routing mechanism,
The routing mechanism changes the routing table from the optimum route to the detour route, and returns the routing table to the optimal route when the detour condition disappearance information is sent. In the detour method of the packet switching system, n times within a predetermined Ta ₁ time. When the above no response occurs, or when the packet is held invalid even after a predetermined Tb ₂ hours have elapsed since it was connected to the transmission queue corresponding to the line,
Assuming that the quality of the transmission line is temporarily degraded, a detour request due to non-response occurrence or a detour request due to invalid hold is sent, and a predetermined Ta ₂
After a lapse of time or a predetermined Tb ₃ hours after sending the bypass request due to the invalid hold, a transmission path quality monitoring means (36) for sending a bypass cancellation request to the routing mechanism is provided, and the bypass request is transmitted by the routing mechanism. The detouring method is characterized in that the routing table is changed from the optimum route to the detour route when the detour cancellation request is sent, and the routing table is returned to the optimal route when the detour cancellation request is sent.