JPS6310934A - Packet communication control equipment - Google Patents

Abstract

PURPOSE:To earlier find out and to restore a call establishing state left ineffective due to a fault of a PT of a caller side of a called party side by checking the effectiveness of each call periodically so long as there is a call established in a public packet exchange network via a private packet exchange network. CONSTITUTION:When a call request packet CR exists from a PT 105B in a logical channel N establishing a call as a called party, it is not a normal procedure. For example, it means that the PT 105B handles the call to be cut off by its own disposal depending on its internal processing and recognizes or treats a logical channel N as an idle channel. On the other hand, the logical channel N is left as call establishing state in the system. Thus, a call request packet CR is stored once in a memory to enter the cut-off of ineffective call establishing. That is, a restoration request packet CQ is generated in the inside of a private DSU 104B and sent to a node 103B. Thus, a GWS 101, a PSDN 100 cuts off/restores an ineffective call.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a packet communication control device, and particularly to a device for effectively disconnecting packet-level logical links that are left invalid between a private packet network and a public packet network. The present invention relates to a packet communication control device that has been improved as much as possible.

[Background Art] In the era of advanced information and communication, various methods for constructing information and communication systems have been proposed. Among them, a system that connects dispersed business offices with a public packet switching network and installs a private packet switching network connected to the public packet switching network within each business office is attracting attention. This system is highly flexible in that it is possible to expand the single logical multiple packet interface of the public packet switching network to multiple logical multiple packet interfaces, and to distribute these interfaces within a business office. It can be said to be a method.

FIG. 5 is a block diagram showing an example of a system in which offices A and B are connected by a public packet switching network and a private packet switching network. That is, two offices A.

8 are connected via a public packet switching network, and a local area network (LAN) is installed within each office. In the figure, 100 is a public packet switching network (P
SDN), 101 is the gateway server (GWs), 1
02 is a LAN cable, 103 is a LAN communication control device (
node).

104 is X, 25 communication control unit (private DSU), 105
is an X,25 packet type terminal (PT). still,
PSDN-GWS and private DS in this system
The interface between U-PT is based on the International Telegraph and Telephone Consultative Committee (CC) at the physical level, link level, and packet level.
ITT Recommendation x, 25 is fully complied with.

Here, the private DSU 104 provides the same interface as the Digital Service Unit (DSU) in PSDNloo. Therefore, PTI05 is private DSU1
04 is connected to the above three levels. Each PT105
The address assigned to this LAN system is
When joining DNloo, the subscriber line number (DTE number) given by the PSDN side and each
It consists of an extension number for identifying T105.

The extension number is generally the node address that connects the PT 105 to the LAN, and the calling PT 105 uses the extension numbers of the PT 105 and the called PT 105 in the call user data section of the calling packet and the calling packet. added to.

The GWS 101 is a device that matches the interface between the PSDNloo and the LAN system, such as protocols and transfer speeds, and from the perspective of the PSDNloo, it becomes a data terminal equipment (DTE) having a plurality of logical channels. Each logical channel has G during the packet-level connection sequence.
Dynamically allocated by WS 101 and PSDNloo. Note that since the nodes are already known, their explanation will be omitted.

FIG. 6 is a diagram illustrating a typical packet sequence from call establishment to recovery. In the diagram, 105A is PT of office A, 105B is PT of office B, and 200 is PT105.
This is a collection of networks between A and 105B. Since this packet control effort is already known, the explanation will be omitted.

FIG. 7 is a diagram illustrating the flow of packets from a calling packet CR to an incoming packet CN. Calling side PT 10
The calling packet CR sent from 5A is private DSU10
At 4A, the link level header is removed and the packet is CR.
Become I. The packet CRI is added with a LAN header, becomes a packet CR2, and is sent to the GWS 101A. In GWSIOIA, the LAN header is removed and a link level header is added, resulting in a packet CR.

GWSIOIA uses a free logical channel to send packet C.
Send R to PSDNloo. The packet CR routed within the PSDNloo by the called DTE number in the packet CN is converted into a called packet CN within the PSDNloo, and is received on the empty logical channel of the called side GWS 101B. Furthermore, the GW3101B removes the link level header, reads the extension number of the called party from the call user data field in the packet CN, adds a LAN header with that number as the reception address, and sets the packet to CN2. Send up.

The packet CN2 received by the node 103B has its LAN header removed and becomes a packet CNI, which is sent to the private DSU1.
Passed to 04B. The private DSUI04B adds a link level header to the packet CN, and then sends it to the called side PT1.
Give it to 05B. The above is the flow of packet transfer.

Note that the destination of the calling packet CR is Pr2O within the same LAN.
3, the received private DSU104 converts it into a packet CN. Whether the communication is within the same LAN can be determined from the fact that the DTE numbers of the calling and called sides are the same.

As mentioned above, this system can connect multiple private DSUs to the LAN by connecting PSDNloo's single DSU.
Although this is an excellent system that can be extended to the SU 104 and enables intra-LAN communication, it has the following drawbacks.

During the data transfer phase shown in FIG. 6, GWSlol, PSDNloo,
In both cases, it is impossible to know the abnormality on the DSUIO4 side. Therefore, calls to the logical channel in use that has been established between the GWSIOI and PSDN 100 are not disconnected or restored, and the logical channel in use is left unused, which wastes communication resources on both the GWSIOI and PSDN 100. The disadvantage is that it occurs.

There are two conventional measures. One is private DSU104
When the side returns to normal state, a restart request packet is sent to the GWS 101, and the GWS 101
PSD disconnects and restores all logical channels assigned to
The countermeasure is to act on behalf of Nloo. but,
With this measure, there is no guarantee that logical channels are disconnected and restored at an early stage.

Another measure is to wait for a disconnection request from the called PT 105. This means that if the transmitted data from the calling side PTI05 cannot be received for a period longer than the specified time determined by the attributes of each PT105, the called side PT105 determines that there is a transmission error and disconnects the call within the PT105. It takes advantage of this fact. In this case, the called side PT1
If 05 sends a recovery request packet to the calling PT 105, both the GWSIOI and PSDN 100 will disconnect the call.
Recovery can be performed. However, the called PT
As a function of 105, it is not specified that a recovery request packet is sent to PSDNloo when there is an abnormal reception. Therefore, the network side cannot expect such behavior from the called side PT 105.

[Problems to be Solved by the Invention] The present invention has been made in view of the above-mentioned drawbacks of the prior art. An object of the present invention is to provide a packet communication control device capable of early discovering, effectively disconnecting and restoring a call that is currently in use and has been left invalid.

[Means for Solving the Problems] In order to solve the above problems, the packet communication control device of the present invention includes verification means for verifying the validity of the call established for each logical channel at the packet level.

[Operation] In such a configuration, the validity of the call establishment state in the system connecting the private packet switching network to the public packet switching network can be verified as long as there is a logical channel establishing a call via the packet communication control device. , for example, periodically by energizing timer means.

[Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram of the X, 25 packet communication control device (private DS0104B) of the embodiment. Private DSU
104B replaces the conventional DSU 104B in the network system of FIG.

In the figure, 102B is a LAN network cable,
103B is a LAN communication control device (node), 104B is a private DSU of the embodiment, and 105 is an x, 25 packet type terminal device (PT105B).

In private DSt1104B, 401 is node 103
Node interface for data communication with B, 4
02 is a state variable management table that stores sequence numbers of transmitted and received frames, etc., 403 is a system constant management table that stores maximum packet length, timer value for transmitting and receiving confirmation, etc., and 404 is an address management table that stores DTE numbers, extension numbers, etc. , 405 is a calling side logical channel management table used for registering/updating the logical channel number called from the PT 105B, and 406 is an incoming call table used for registering/updating the logical channel number assigned to the packet that arrives addressed to the PT 105B. side logical channel management table, 407 is a C line bell monitoring circuit that monitors the ready state of PT105B, 408 is a line speed clock generation circuit that supplies a transmission/reception clock to PT105B, 409 is a terminal interface for data communication with PT105B, 410 is a test loop circuit that forms a loopback of communication data;
411 is a loop switch (
SW), 412 is a high-level data link control unit (HDLC) that performs data transmission and reception in frame units.
), 413 is a direct memory access (DMA) control unit that controls data transfer between the PT 105B and the memory 414 and between the memory 414 and the node 103B, 414 is a memory circuit that stores programs, transmitted and received data, etc., 415
416 is a control unit (CPU) that controls the private DSU 104B, decodes packets, disassembles packets, and controls the timing of communication operations.

[First Embodiment] FIG. 2 is a flowchart showing the packet communication control procedure of the first embodiment performed by the CPU 416.

In the first embodiment, when a flow of packets that cannot occur in a normal communication procedure occurs in a logical channel (N) in operation, some kind of failure causes the flow of packets in the logical channel (N).
) is left inactive, and the call is immediately disconnected and restored. - As an example, called party D
Logical channel (N) establishing call for SU 104B
In contrast, the case where there is a call request CR from P7105B will be described.

In steps 31 to S5, an incoming call reception packet is detected. That is, in step S1, it waits for packet reception from the PT 105B, and when a packet is received, the link level header is deleted in step S2. In step S3, it is determined whether the packet is an incoming call acceptance packet CA or not, and if it is other than the incoming call acceptance packet CA, it is sent as is to the node 103B in step S15.

Further, if it is determined in step S3 that the incoming call acceptance packet CA is received, the PT 105B acts as the called PT by other PTs 105A (or P7105 in the same LAN 102B) in the system.
It is determined that the call has been established with the call destination (or B'), and the logical channel (N) on which the call was established is registered in the called side logical channel management table 406 in step S4. Step S5
Then, the incoming call acceptance packet CA is transmitted to the node 103B.

In steps S6 to SIO, the usage status of the logical channel (N) that established the call is monitored. That is, if it is determined in step S8 that the received packet from the PT 105B uses a channel other than logical channel (N), step S1
8, the packet is sent as is to the node 103B. If it is determined in step S8 that the received packet is for the logical channel (N), the process advances to step S9 to check whether it is a disconnection confirmation packet CF. If it is a disconnection confirmation packet CF, it is determined that the call on the logical channel (N) has been disconnected,
In step S16, the called side logical channel management table 40
6, the call establishment for logical channel (N) is deleted. In step S17, a disconnection confirmation packet CF is transmitted to the node 103B. The flow returns to step S1 to detect call establishment again.

Further, if it is determined in step S9 that the packet is not a disconnection confirmation packet CF, it is checked in step SIO whether it is a call request packet CR, and if it is not a call request packet CR, the process advances to step S18. However, when it is determined that the packet is a call request packet CR, it is considered that some kind of abnormality has occurred. That is, if there is a call request packet CR from the PT 105B on the logical channel (N) on which the call has been established as the called party, this cannot be said to be a normal procedure. For example, the PT 105B automatically disconnects the call in its internal processing, and recognizes or treats the logical channel (N) as an empty channel. On the other hand, in the system, the logical channel (N) was left in the call establishment state. Therefore, in step Sll, the call request packet CR is temporarily stored in the memory, and an operation for disconnecting the invalid call establishment is started. That is, in step S12, the restoration request packet CQ of the logical channel (N) is sent to the private DSU 10.
4B and sends it to the node 103B in step S13. This allows GWS 101, PSDNlo
o can disconnect and restore invalid calls. In step S14, the call request packet CR stored in the memory is
is transmitted to the node 103B, and the process returns to step S1.

The above is grinning, private DSU104B is GWSIOl, PS
An invalid call left on the DN 100 can be detected, and the DSU 104B itself can detect the recovery request packet C.
By creating and sending Q, automatic disconnection and recovery of invalid calls becomes possible.

[Second Embodiment] FIGS. 3(a) and 3(b) are flowcharts showing the packet communication control procedure of the second embodiment performed by the CPU 416.

In the first embodiment, the existence of an invalid call is detected when the PT 105B makes an abnormal call request. On the other hand, in the second embodiment, the private DSUIO 4B periodically transmits a pseudo-incoming call packet CN' to the PT 105B, thereby monitoring the response of the PT 105B, and detecting the abandonment of an invalid call at an early stage.

P71 in steps S31 to S33 and step S53
The call acceptance packet CA and the disconnection confirmation packet CF are found from the packet transmitted in 05B, and the call establishment and disconnection are updated in the called side logical channel management table 406.

That is, when it is determined in step S32 that it is an incoming call acceptance packet CA, the logical channel (N) is registered, and the process proceeds to step S32.
When it is determined in step 53 that the packet is a disconnection confirmation packet CF, the registration of the logical channel (N) is deleted. In step S34, the received packet CA or CF is transmitted to the node 1o3-B. Furthermore, in step S35, it is checked whether there is a logical channel on which a call has been established in the called side logical channel management table 406, and if there is, a timer 2 is activated in step S36. Timer I is a timer for detecting whether the logical channel (N) in the call established state is left invalid for a predetermined period of time in the private DSU 104B.

The value of the predetermined time is, for example, longer than the time during which the normal packet sequence shown in FIG. 6 is performed, and depends on the maximum amount of data transfer. Further, when there is no logical channel in the call established state, the timer (2) is stopped/reset in step S54 so that unnecessary timer interruptions do not occur.

When timer I times up in step S37, call inspection is started. First, in step 538, timer I is stopped, and timer I is initialized to start the next timer.

In step S39, the called side logical channel management table 4
Select one logical channel (N) in call established state from 06. In step S40, a pseudo incoming call packet CN' for the logical channel (N) is created in the private DSU 104B,
Add link level header. In step S41, the pseudo call packet CN= is transmitted to the PT 105B. In step S42, timer II is activated. Timer II is a timer that waits for a response from the PT 105B in response to the transmission of the pseudo-incoming call packet CN'. The time-up time of the timer 11 is determined with sufficient margin in consideration of the normal response time from the PT 105B. Therefore, if the call acceptance packet CA for the logical channel (N) is not returned by the time the timer II times up, it is considered that the pseudo call packet CN' has been discarded within the PT 105B. That is, it can be determined that the logical channel (N) is currently being used within the PT 105B.

To explain the above specifically, in step S43, P71
Waits for a received packet from 05B, and when received, proceeds to steps S44 and 45 to check whether the logical channel number of the received packet is (N). If the logical channel number is (N
), the process proceeds to steps 358 to S61,
If necessary, the called side logical channel management table 405 is updated and sent to the node 103B. Furthermore, even if it is determined in step S45 that it is a logical channel (N), if it is determined in step S46 that it is not the incoming call acceptance packet CA, the process proceeds to step S60. This is so that it does not interfere with the normal use of the logical channel (N).

When the logical channel (N) is determined in step S45 and the incoming call acceptance packet CA is determined in step S46,
The received packet CA is considered to be a response to the pseudo-incoming call packet CN' sent in step S41.

Moreover, the fact that there is a response to the incoming call acceptance packet CA at this point means that the logical channel (N) is PT105B.
This means that the call was being treated as disconnected internally before the pseudo-incoming call packet CN' was received. That is, the logical channel (N) was left in an invalid call established state. The flow advances to step S47 to stop/reset timer II. In step S48, private DSU1
Recovery request packet CQ of logical channel (N) in 04B
is created and transmitted to the node 103B in step S49. As a result, the GWSIOI and PSDNloo restore the invalid call establishment state of the logical channel (N). In step S50, the private DSU 104B creates a disconnection instruction packet CI for the logical channel (N), and in step S51, the private DSU 104B creates a disconnection instruction packet CI for the logical channel (N).
105B. This is the pseudo incoming call packet CN'
Since the PT105B that responded is in the data transfer phase state, the PT105B is responsible for the private DSU104B.
This is to restore the 5B logical channel (N). In step S52, the called side logical channel management table 40
The call establishment for the logical channel (N) of No. 6 is deleted, and the process proceeds to step S57. In step S57, a check is made to see if all the logical channels (N) in the call established state in the called side logical channel management table 406 have been checked using the pseudo call packet CN'', and if the check has not been completed yet, step Return to S39, and if finished, return to the start.

Also, in step S55, the logical channel l is transferred from the PT 105B.
), when timer II times up before the incoming call acceptance packet CA arrives, the pseudo incoming call packet CN' is PT1.
It is thought that it was discarded within 05B. That is, PT10
It can be determined that the logical channel (N) inside 5B was already in an operating state in the data transfer phase. The flow advances to step S56, where timer II is stopped and reset to the initial state. In step S57, the existence of the next logical channel (N) to be tested is checked.

As described above, according to the second embodiment, the logical channel (N) registered in the call established state is periodically checked, and if there is a call that has been left invalid, it can be detected early. GWSI from private DSU104B
It becomes possible to recover invalid calls within the OI and PSDN 100.

[Third Embodiment] FIGS. 4(a) to 4(c) are flowcharts showing the packet communication control procedure of the third embodiment performed by the CPU 416.

In the second embodiment, the logical channels are periodically checked by the timer I, which is started only when there is a logical channel (N) in the process of establishing a call.

On the other hand, in the third embodiment, when an incoming call packet CN based on a call request CR from another PT 105A or the like in the system arrives at the PT 105B, the call is checked using the incoming call packet CN. It is.

When receiving call acceptance packet CA from P7105B (
Step 5 IOI (5102), step 5103 registers the call establishment in the called side logical channel management table 406. In addition, a disconnection instruction packet CI (
Upon receiving a recovery request (steps S71, 572),
In step S93, the called side logical channel management table 40
6. Delete the call establishment registration. In this way, the latest call establishment status is constantly stored in the table 406.

In step S73, the incoming packet CN from the node 103B
When the incoming call packet CN is received, in step S74, the incoming call packet CN
is stored, and in step S75 it is determined whether or not there is a logical channel (N) in the call established state in the table 406. If it does not exist, proceed to step 5108 and PT105
Check whether there is a free logical channel to B. If there is a free logical channel, use that logical channel to perform PTIO.
Send the incoming call packet CN to 5B (steps 5l11 to 5
113). Also, if there are no free logical channels, step 5
109, the incoming call packet CN is a recovery request packet CQ.
and is transmitted to the node 103B in step 5itO. In other words, the incoming call will be rejected.

Now, if it is determined in step S75 that there is a logical channel (N) in use in the table 406, the process advances to step 376 and a call check is started. This checking method uses the incoming call packet CN stored in step S74 to create a pseudo incoming call packet CN' in which it is assigned to the logical channel (N) in which the call is currently being established.
By sending a pseudo incoming call packet CN' to B,
Depending on how the PT105B responds, the logical channel (
This is to judge the effectiveness of N).

Specifically, in step 376, a logical channel (N) in use is selected from the called side logical channel management table 406, and in step S77, a pseudo An incoming call packet CN' is formed and sent to PT 105B in step 378.

In step S79, timer II is activated. Timer II
The function of is the same as that of timer II in the second embodiment.

In step s80, a received packet from the PT 105B is waited for, and when it is received, the process advances to steps S81 and 82 to check whether the logical channel number is (N) or not. If the logical channel number is not (N), the received packet is transmitted to the node 103B in step S96. Also, step S82
Even if it is determined in step S83 that it is a logical channel (N), if it is determined in step S83 that it is not the incoming call acceptance packet CA, the process advances to step 396. This is so that it does not interfere with the normal use of the logical channel (N).

When the logical channel (N) is determined in step S82 and the incoming call acceptance packet CA is determined in step S83,
The received packet CA is sent to the PT 105 in step 378.
This is considered to be a response to the pseudo-incoming call packet CN' sent to B. Furthermore, since there is a response to the incoming call acceptance packet CA at this point, the CPU 416 causes the current channel (N) to be treated as a call disconnection within the PT 105B even before receiving the pseudo incoming call packet CN'. It can be concluded that it was. The logical channel (N) was left in an invalid call establishment state. The flow is step S8
Proceed to step 4 to stop/reset timer II. In step S85, the incoming call acceptance packet CA is stored, and in step S86, the logical channel (N
) is created and transmitted to the node 103B in step S87. This allows GWS 10
1.

PSDNloo restores the invalid call establishment state of logical channel (N).

In step S88, it waits for packet reception from node 103B. When a packet is received, it is checked in step S89 whether the logical channel number is (N). Logical channel (N)
If so, it is checked in step S90 whether it is a recovery confirmation packet CF. If the recovery confirmation packet CF, step S85
The incoming call acceptance packet CA stored in is transmitted to the node 103B. That is, as a result of the inspection using the regular incoming call packet cN received in step S73, the private DsU 104B detected and released the logical channel (N) that had been left invalid, so the above-mentioned pseudo incoming call packet C.N.
' Logical channel (N) connected with PT105B
In order to utilize the link state as is, the call acceptance packet CA stored in step S85 is sent to the other party's PT 105A. In step S92, the call establishment of the logical channel (N) is registered in the called side logical channel management table 406.

If it is not a logical channel (N) in step S89, it is determined in step S97 whether or not it is a disconnection instruction packet CI.

If it is a disconnection instruction packet CI, the corresponding channel in the called side logical channel management table 4.6 is updated in step S98. In step S99, a link level header is added, and in step 5100, the disconnection instruction packet CI is added to the PT105.
Send to B. Further, if it is determined in step S90 that the recovery confirmation packet is not CF, the process advances to step s99.

Further, if the timer II times up before the incoming call acceptance packet CA for the logical channel (N) arrives from the PT 105B in step 5105, it is considered that the pseudo incoming call packet CN' is discarded within the PT 105B. That is, within the PT 105B, the logical channel (N) is determined to be in an operating state in the data transfer phase.

Flow proceeds to step 5106 where Timer II is stopped and reset to the initial state. In step 5107, it is checked whether all logical channels (N) have been checked. If checked, there was no logical channel in an invalid call established state, so the process goes to step 5108, and a new logical channel is allocated to the logical channel. Find logical channels.

[Effects of the Invention] As explained above, according to the present invention, as long as there are calls established on the public packet switching network via the private packet switching network, the validity of each call is periodically checked. A call establishment state that has been left invalid due to an abnormality on the calling side or the called side PT can be discovered and restored at an early stage, and therefore the waste of communication resources can be minimized.

[Brief explanation of drawings]

Figure 1 shows the packet communication control device (private DSU) of the embodiment.
FIG. 2 is a flowchart showing the packet communication control procedure of the first embodiment, FIGS. 3(a) and (b) are flowcharts of the packet communication control procedure of the second embodiment, and FIG. ) to (C) are flowcharts showing the packet communication control procedure of the third embodiment, and FIG. 5 is a block diagram showing an example of a system connecting offices A and B with a public packet switching network and a private packet switching network. , FIG. 6 is a diagram showing a general packet sequence from call establishment to recovery, and FIG. 7 is a diagram explaining the flow of packets from calling packet CR to called packet CN. In the figure, 102...LAN network cable, 10
3... LAN communication control unit (node), 104...
Private DSU, 105... , 405...Calling side logical channel management table, 406...Calling side logical channel management table, 40
7...C line bell monitoring circuit, 408...Line speed clock generation circuit, 409...Terminal interface, 4
10... Test loop circuit, 411... Loop switch (SW), 412... High level data link control control unit (HDLC), 413... Direct memory access (DMA) control unit, 414...・
・Memory circuit, 415...Timer circuit, 416...
This is a control unit (CPU). Patent applicant: Canon Co., Ltd. Figure 4 (C) Figure 6

Claims (1)

[Claims] In a packet communication control device that controls packet communication between the private packet switching network and packet terminals in a system in which a private packet switching network is connected to a public packet switching network, the effectiveness of the call established for each logical channel is measured at the packet level. A packet comprising: a verification means for verifying call validity by the verification means, and verification of the validity of the call by the verification means is performed periodically as long as a logical channel establishing a call via the packet communication control device exists. Communication control device.