JP3046422B2 - Connectionless communication method in ATM network - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention accommodates local connectionless information, such as LAN data of a local area network, in an asynchronous transfer mode (ATM) network using a connection-oriented communication system, and comprises a plurality of LANs.
The present invention relates to a connectionless communication system in an ATM network for interconnecting the two.

[0002]

2. Description of the Related Art There is an increasing demand for connection between LANs by accommodating local connectionless information such as LAN data in a global connection-oriented ATM network.

[0003] The connectionless information is generally of variable length, and a destination address is written at the head thereof. In order to accommodate such connectionless information in an ATM network that exchanges information as fixed-length cells, it is necessary to perform cellization / restoration of variable-length data, destination analysis for each cell, and routing control.

Generally, in an ATM network accommodating connectionless information, a connectionless service function (CLSF) for performing address analysis and cell routing control for each message, that is, a connectionless communication server is provided in view of its processing capability. From within the network. In this case, how to connect the servers becomes a problem.

Conventionally, as a server connection method, a mesh connection method and a tree connection method shown in FIGS. 15A and 15B are known. The mesh connection method has the advantage that the signal delay is small, but when adding a server, it is necessary to change the routing tables in all the servers, and furthermore, in order to determine to which server the message is sent, the message It is necessary to analyze many digits of the address, and there is a problem that it is not suitable for a large-scale network. The tree connection method has an advantage that the amount of the routing table is small, but has a problem that the relay server is likely to be overloaded because all traffic between the stations passes through the relay server.

For this reason, the applicant of the present invention has proposed a connectionless communication system using an ATM network which can perform routing efficiently and at high speed and can reduce a load on a server. FIG. 16 is a block diagram of such a proposed connectionless communication system. Reference numeral 1 denotes a connectionless information celling / restoring unit, which has variable-length connectionless information such as LAN data and fixed-length cells used in an ATM network. It converts data from the LAN into a plurality of fixed-length connectionless cells and outputs the data. Reference numeral 2 denotes a connectionless communication server installed in an exchange in the ATM network, which analyzes, for example, a destination address in a connectionless cell converted from LAN data and controls the routing of the cell in the ATM network. 3 is for transferring fixed-length cells in the asynchronous transfer mode.
A TM network which connects the connectionless information celling / restoring unit 1 and the server 2 by a fixed path, for example, a permanent virtual circuit, and connects the servers 2 with each other as a fixed path, such as a permanent virtual circuit or a semi-fixed path, for example. Connect by virtual circuit.

The connectionless information (message) as variable-length LAN data is divided into a plurality of fixed-length cells by a connectionless information celling / restoring unit 1 and the same message identifier MID is added to each cell. I do. The connectionless information celling / restoring unit 1
BOM (beginning of message) is added as the segment type to the first cell of the divided message, and EOM (end of message) is added to the last cell.
, And COM (continuation of message) to all cells in the middle, and SSM (single segment message) when one message is converted to only a single cell. The server 2 searches for routing information in the ATM network from the destination address of the connectionless information included in the cell to which the BOM or SSM is added, and transfers the connectionless information to the destination server via the path indicated by the routing information. . Server 2
Stores the retrieved routing information in a built-in storage unit in correspondence with the message identifier MID of the cell to which BOM or SSM is added.

[0008] Thereafter, when cells to which COM or EOM is added are input, the server 2 searches the storage unit for routing information corresponding to the MID of these cells, and performs connectionless connection via the path indicated by the routing information. Transfer the information to the destination server. Then, after transferring the last cell of the message, that is, the cell to which EOM or SSM is added, to the destination server via a predetermined path, the correspondence between the MID and the routing information is deleted from the storage unit, and one connectionless The communication of information ends.

FIG. 17 is a block diagram of a connectionless communication server. A segment type separation unit 21 separates a segment type from an input connectionless cell.
Detect its contents. The detected segment type is B
In the case of OM or SSM, the routing information search unit 22 analyzes the destination address of the LAN data included in the cell, searches for routing information in the ATM network, and outputs the contents via the OR circuit 23. Output to the routing information adding unit 24. Routing information adding unit 24
Outputs the cell added with the routing information sent from the routing information search unit 22 to the cell to which BOM or SSM is added as the segment type.

[0010] The routing information searched for the cell to which the BOM or SSM is added is temporarily stored in the MID / routing information temporary storage unit 25 in association with the message identifier MID added to the cell. When a cell whose segment type is COM or EOM is input, the routing information is searched based on the correspondence between the MID and the routing information temporarily stored in the MID / routing information temporary storage unit 25, and the information is searched. Is supplied to the routing information adding unit 24 via the OR circuit 23, and the routing information is added to the cell.

[0011] When a cell to which EOM or SSM is added as a segment type is input, routing of the cell of the message is terminated.
The ID erasure unit 26 outputs an erasure signal to the MID / routing information temporary storage unit 25, and the correspondence between the stored message identifier MID and the routing information is erased.

As described above, one message as connectionless information is divided into a plurality of cells, the routing information of the cell is searched using the destination address of the message included in the first cell, and the cell in the middle is searched. For the last cell, the same routing information is searched by the message identifier, and routing is performed.

FIG. 18 shows a connection system between servers adopting the above connectionless communication system. In each hierarchy, a plurality of servers are grouped by a plurality of groups, mesh-connected in the group, and upper hierarchy. Connected to the relay server. In the server connection method of this hierarchical structure, the 10-digit address _{X 1, X 2, Y 1} , Y 2, Z 1, Z 2,
Of A ₁ , A ₂ , A ₃ , and A ₄ , 6 of X ₁ X ₂ Y ₁ Y ₂ Z ₁ Z ₂
Each layer is configured to correspond to an address of two digits, and the server of each layer analyzes the destination address of the message from the upper digit, and if the address is other than the own group, the upper layer The message is routed to the relay server. If the message is addressed to the own group, the lower digit is analyzed, and the message is routed to the server on the other side.

[0014]

In the above connectionless communication system, the message identifier MID added to the connectionless cell is represented by 10 bits.
For this reason, the server needs an area for storing 2 ¹⁰ (= 1024) types of information for each process per subscriber (one LAN). For example, the MID / routing information temporary storage unit in the server has 2 ¹⁰ (= 102
4) A storage area corresponding to the type of MID is provided, and routing information is stored in a storage area corresponding to the MID of the input headless connectionless cell. For this reason,
A memory for storing a huge amount of information of 1024 types is required for each subscriber, and there is a problem that the cost increases.

As described above, an object of the present invention is to provide a connectionless communication system in an ATM network in which a storage area in a server can be reduced. Another object of the present invention is to provide an A which can easily identify an error message and can easily discard cells belonging to the error message.
An object of the present invention is to provide a connectionless communication system in a TM network. Another object of the present invention is to provide a connectionless communication system in an ATM network in which the number of simultaneously transmittable / receivable messages can be set for each subscriber.

[0016]

FIG. 1 is a diagram illustrating the principle of the present invention. Reference numeral 11 denotes a plurality of connectionless information celling / restoring units for performing bidirectional conversion between variable-length connectionless information and connectionless cells as fixed-length cells handled in an asynchronous transfer mode (ATM) network;
A plurality of connectionless communication servers (hereinafter simply referred to as servers) provided in a switch of the network for analyzing a destination address in the connectionless cell and performing routing of the connectionless cell, and 13 via a path indicated by the routing information It is an ATM network that connects servers. At the server, 21 is the message identifier M of the cell
M to convert ID to server message identifier SMID
An ID / SMID conversion unit, 22 is a message processing unit that performs message processing such as routing information search based on the SMID, and 23 is an S / D obtained by the MID / SMID conversion unit.
An SMID notifying unit that notifies the message processing unit of the MID.

[0017]

[Operation] Connectionless information celling / restoring unit 11
Divides one message as variable-length connectionless information into a plurality of fixed-length cells, adds the same message identifier MID to each cell, and assigns the first cell of the divided message as a segment type B
OM, EOM in the last cell, C in all cells in the middle
OM, SSM is added when one message is converted to only a single cell, and each server is connected to the originating server 1 via a permanent virtual circuit as a fixed path.
Enter 2

The server 12 is set with a smaller number of in-server message identifiers SMID than the message identifier MID, and the MID / SMID conversion unit 21 sets the message identifier MID of the cell to an unassigned value when a BOM cell or SSM cell is input. The message processing unit 22 converts the SMID to the used SMID, determines that the SMID is in use, temporarily stores the correspondence between the MID and the SMID, and searches the routing information in the ATM network from the destination address DA included in the BOM cell or the SSM cell. Sends the routing information to the destination server via a path indicated by the routing information, and sends the routing information to the SMI.
The data is temporarily stored in association with D.

In this state, when a COM cell or an EOM cell is input, the server 12 changes the MID of the cell to the MID.
Converted to SMID using the correspondence between ID and SMID,
The routing information according to the obtained SMID is stored in the SM.
It is obtained from the correspondence between the ID and the routing information, and is transmitted to the destination server in connectionless cell units via the path indicated by the routing information.

As described above, since it is only necessary to store the MID and the routing information in correspondence with the in-server message identifiers SMID, which is smaller than the message identifier MID, the storage area in the server can be reduced.

In particular, since the number of messages that can be transmitted and received simultaneously per subscriber is much smaller than the types of MIDs (2 ¹⁰ = 1024 types), the number of SMIDs per subscriber is the number of messages that can be transmitted and received simultaneously. As a result, the storage area in the server can be reduced, and the SMID can be assigned to all the messages in service. By limiting the number of SMIDs, it is possible to limit the number of messages that can be transmitted and received simultaneously per subscriber.

Furthermore, when a cell to which EOM or SSM is added is inputted, each temporarily stored correspondence is erased, so that only the correspondence corresponding to the message in service at all times can be stored, and the memory capacity can be reduced. Can be reduced.

The SMID notifying section 23 outputs the MID
If the SMID obtained by the SMID conversion unit 21 is added to an empty area of each cell or an area not used in the subsequent processing and notified to the message processing unit, MID / S
Physical signal lines between the MID conversion unit and the message processing unit can be reduced.

Further, the MID / SMID conversion unit 21 stores the use / unused state in correspondence with each SMID, and assigns an unused SMID to the cell when a BOM cell or an SSM cell is input. During use, and E
The SMID management unit that releases the SMID by not using it when the processing of the MO cell is completed, stores the correspondence between the MID and the SMID, and uses the correspondence between the MID and the SMID when a COM cell or an EOM cell is input. The MID registration / conversion unit that converts the MID of the cell into the SMID and a cell discard unit that discards the cell when a cell having an MID not registered in the MID registration / conversion unit arrives , The SMID can be easily captured and released for each message, and the cell belonging to the abnormal message can be easily detected and discarded.

Also, whether each SMID is used or not used,
If the correspondence between D and MID is stored together,
Conversion from MID to SMID and management of SMID can be physically realized by one mechanism, and the storage capacity can be reduced at the same time.

[0026]

EXAMPLES Overall configuration FIG. 2 is a block diagram illustrating an overall structure of a communication system that can be applied connectionless communication method of the present invention. In the figure, reference numeral 11 denotes a connectionless information cell forming / restoring unit for performing bidirectional conversion between LAN data as variable-length connectionless information and connectionless cells (ATM cells) as fixed-length cells handled in an ATM network; Reference numeral 12 denotes a plurality of servers provided in the exchange of the ATM network for performing routing, screening of incoming / outgoing calls, and flow control. Reference numeral 13 denotes an ATM network for transferring fixed-length cells in an asynchronous transfer mode. Unit 11 and server 12
The servers 12 are connected by a fixed path such as a permanent virtual circuit, and the servers 12 are connected by a permanent virtual circuit or a semi-fixed path such as a virtual circuit as a fixed path.

When variable-length LAN data, ie, a message, is sent from the calling terminal 16 among the terminals on the calling LAN,
The originating connectionless information celling / restoring unit 11 divides the message into fixed-length cells that can be handled by the ATM network, adds the same message identifier MID to each cell, and adds the first cell of the divided message. , BOM as the segment type, EOM in the last cell, COM in all the cells in the middle, and SSM when one message is converted into only a single cell. Note that the destination information of the message is included only in cells to which BOM or SSM is added (referred to as BOM cells and SSM cells).
OM or EOM-added cell (COM cell, EOM
Does not include destination information.

Each of these divided cells has an ATM network 1
The ATM header required for the exchange in the communication unit 3 is added, but the connectionless information celling / restoring unit 11 on the calling side is added.
Then, only the virtual identifier for identifying the fixed path to the originating server 12 is added to the ATM header without analyzing the destination of the message.

Thus, the input message (LA)
N data), the connectionless cells obtained by dividing the data reach the originating server 12 and traverse the ATM network 13 based on the routing control by the server 12 described later.
2 is transferred to the connection-less information celling / restoring unit 11 on the receiving side via a fixed path.
The TM cell is converted again into variable-length LAN data and transferred to a predetermined destination terminal 18 among destination terminals.

Although only two servers 12 are shown in FIG. 2, when a wide-area ATM network is used, a plurality of servers are used and the two servers are connected based on destination information of LAN data. ATM by repeating routing
Routing throughout the network can be performed.

FIG. 3 shows an embodiment of a format of an ATM cell used in the present invention. In the figure, an ATM cell has a fixed length of 53 bytes, of which 5 bytes are ATM bytes.
The header, 48 bytes, is used as an information field. The 48-byte information field includes a 2-byte header used for routing connectionless cells, 44-byte connectionless data, and a 2-byte trailer in which a payload length and the like are stored.

In the ATM header, GFC is a 4-bit generic flow control used for flow control between links, VPI is an 8-bit virtual path identifier,
VCI is a 16-bit virtual channel identifier, PT is a payload type indicating a cell type, RA is a reserve bit, and PR is 1 used for cell discard control at the time of congestion.
The cell priority of bits, HEC, is an 8-bit header control for correcting and detecting bit errors.

ST is a segment type in the adaptation field except for 44 bytes of connectionless data in the information field. The segment type ST is, for example, a layer 3 protocol data unit (L3-PD) as LAN data.
When U) is divided into a plurality of ATM cells, it indicates the position of the cell, and indicates the above-mentioned BOM, COM, EO indicating the head cell, intermediate cell, last cell, and single segment cell.
The distinction between M and SSM is expressed by 2 bits. SN is layer 3
A 4-bit sequence number that is cyclically assigned to a cell in a protocol data unit unit and detects a discard in a cell unit, MID is a 10-bit message identifier, and each AT obtained by dividing one message
The same MID is added to the M cells. PL is a payload length, and CRC is a CRC code.

FIG. 4 is an explanatory diagram of the format of each layer.
A trailer is placed, and the destination address DA is included in the header. The adaptation layer of level 2 divides a variable-length message into 44 bytes each,
A trailer is arranged, and the segment type ST used in the present invention is included in the header as described above.
(BOM, COM, EOM, SSM) and a message identifier MID. A cell in the level 1 ATM layer (ATM cell) has an A at the beginning of the cell in the adaptation layer.
A TM header is arranged, and each ATM header is transferred to a predetermined server via a fixed path indicated by a virtual channel identifier VCI included in the ATM header.

FIG. 5 is a schematic block diagram of the server 12 of the present invention. Reference numeral 21 denotes an MID / SMID conversion unit for converting a message identifier MID of each cell into an in-server message identifier SMID.
Reference numeral 2 denotes a message processing unit that processes a message such as a routing information search based on a destination address or SMID, and 23 denotes an SMID obtained by an MID / SMID conversion unit
Is notified to the message processing unit.
As shown in FIG. 6, the SMID notifying unit 23 adds an SMID to an empty area of a cell or an area not used in subsequent processing (MID area), and an SMID adding unit 23a.
An SMID discriminating unit 23b that discriminates the SMID added by the MID adding unit and notifies the message processing unit 22 is provided. Note that the SMID discriminating unit may be provided in the message processing unit.

The MID / SMID converter 21 has a 10-bit message identifier MID type 2 ¹⁰ (= 102
4) less number of in-server message identifiers SMID
Is set in advance. For example, if the number of messages that can be transmitted and received simultaneously per subscriber is n, n types of in-server message identifiers SMID are set in advance.

In this state, when a BOM cell or an SSM cell is input from the connectionless information celling / restoring unit 11 on the calling side, the MID / SMID conversion unit 21 allocates an unused SMID to the cell. As a result, the cell message identifier MID is converted into an unused SMID, the SMID is used, and the correspondence between the MID and the SMID is temporarily stored.

[0038] The SMID notifying section 23 outputs the MID / SMID.
The SMID obtained by the conversion unit 21 is stored in a BOM cell or SSM
The message processing unit 22 is notified by adding it to an empty area of the cell or an area not used in the subsequent processing.

The message processing unit 22 retrieves routing information in the ATM network from the destination address DA included in the BOM cell or the SSM cell, sends it to the destination server via the path indicated by the routing information, and Information is temporarily stored in association with the SMID.

Next, when a COM cell or an EOM cell is input, the MID / SMID conversion unit 21 converts the MID of the COM cell or the EOM cell into the SMID using the temporarily stored correspondence between the MID and the SMID. I do. SM
Similarly, the ID notifying unit 23 converts the MID / SMID
Is added to the COM cell or the EOM cell and notified to the message processing unit 22. The processing unit 22
The routing information corresponding to the input SMID is obtained from the correspondence between the temporarily stored SMID and the routing information, and is sent to the destination server via the path specified by the routing information.

When the processing for the EOM cell or the SSM cell is completed, the MID / SMID conversion unit 21 deletes the temporarily stored correspondence between the MID and the SMID, and the message processing unit 22 deletes the corresponding SMID from the SMID. The correspondence of the routing information is deleted, and one message communication ends.

FIG. 7 is a configuration diagram of the MID / SMID conversion unit 21. In the figure, reference numeral 21a denotes an SMID management unit, which includes an SMID registration unit 21a-1 and an SMID number control unit 21a-2 for registering / deleting the SMID, as shown in FIG.

The SMID number control unit 21a-2 registers the SMID in the SMID registration unit 21 according to an instruction from a control unit (not shown).
Set (register) or delete to a-1. Note that the number of SMIDs set for one subscriber is the number of messages that can be transmitted and received simultaneously per subscriber (the number of messages that can be simultaneously serviced). In this way, the number of SMIDs set is determined by the type of MID (2
¹⁰ = 1024 types), the storage area can be reduced, and the SMID can be assigned to all messages in service. Conversely, by limiting the number of SMIDs, the number of messages that can be transmitted and received simultaneously per subscriber can be limited.

As shown in FIG. 9 (a), the SMID registration unit 21a-1 stores and manages the use / non-use of each of the set n types of SMIDs, and stores the BOM cells or the SSs.
When an M cell is input, an unused SMID is searched for and output, and the SMID is used, and the EOM is used.
When the processing for the cell or the SSM cell is completed, the corresponding SMID (released SMID) is received and made unused.

Since the server can service a large number of subscribers, if the number of serviceable subscribers is m, mn SMIDs are set. For this reason,
Although not shown, an identification symbol (fixed path name) indicating which SMID is for which subscriber is also stored in the SMID registration unit 21a-1. Then, the SMID number control unit 21a-2 can set and delete the SMID according to the instruction for each subscriber.

Reference numeral 21b denotes an MID registration / conversion unit, and (1)
When a BOM cell or an SSM cell is input, the SMID management unit 2
The correspondence between the SMID output from 1a and the MID of the cell is stored in the registration unit 21b-1 (see FIG. 9B).
(2) When COM cell or EOM cell is input, MID
COM cell or EOM using the correspondence between
The cell MID is converted to SMID and output, and (3) EOM
When processing of a cell or an SSM cell is completed, or when an error occurs due to a format check,
Delete the MID correspondence and release the SMID SM
Notify the SMID management unit 21a as the ID, and (4) the registration unit
When a cell having an MID not registered in 21b-1 is detected, a discard command CNC is output.

Reference numeral 21c denotes a cell discard unit for discarding the corresponding cell in accordance with the discard command CNC. 21d denotes a SMID output from the SMID management unit 21a (when a BOM or SSM cell is input) or a SMID output from the MID registration / conversion unit 21b.
(When a COM or EOM cell is input). Although not shown, the MID / SMID conversion unit 21 separates the segment type ST from the ATM cell, decodes the segment type ST, and
A segment type identification unit for identifying M cells, EOM cells, and SSM cells is included.

In the above description, the SMID management unit 21a and M
Although the registration units 21a-1 and 21b-1 of the ID registration / conversion unit 21b are separately provided, as shown in FIG. 10, the use / non-use of each SMID and the correspondence between the SMID and the MID are integrally stored. It can also be configured as follows. By doing so, MI
Conversion from D to SMID and management of SMID can be physically realized by one mechanism, and the storage capacity can be reduced at the same time. In FIG. 10, the MID column of the unused SMID is set to "1", and the used / unused column is also set to "1"("0" during use).

The operation of the MID / SMID conversion unit 21 will be described below for BOM cell arrival, COM cell arrival, and EOM arrival. Incidentally, the MID and the used / unused are integrally stored in the registration unit 21b-1 as shown in FIG.
It shall be managed.

(A) When a BOM cell arrives (see FIG. 11) When a BOM cell is input, a match search is made to see if an all "1" pattern is registered in the registration unit 21b-1. That is, starting from the head address (head SMID) of the registration unit 21b-1, a match search between the stored contents and the all "1" pattern is performed, and a match is searched for (match search).

The address of the all "1" pattern (SMI
D) if present, the MID of the BOM cell at the address
(10 bits) and data indicating “in use (=“ 0 ”)” are written (written). Next, the SMID, which is the address of the all "1" pattern, is output (SMID output), and the SMID adding unit 23a outputs the SMID to an empty area of the BOM cell or an area (MID area) not used in subsequent processing.
The ID is added and sent to the next message processing unit 22.

If the all "1" pattern does not exist, an outgoing / incoming message has occurred more than the set number of SMIDs, and in such a case, a discard command CNC is output ( Disposal directive). Cell discard unit 21c
When a discard command is issued, the corresponding cell is discarded.

(B) When a COM cell arrives (see FIG. 12) When a COM cell is input, the MID (1
Data DT indicating 0 bit) and being used (= "0") is created, and a search is made to see if a match with the data DT is registered in the registration unit 21b-1. That is, the registration unit 21b-1
Starting from the head address (head SMID), a search is made for a match between the stored contents and the data DT data, and a match is searched for (match search).

If there is a match, the MID of the BOM cell which is the first cell of the message to which the COM cell belongs
(10 bits) has been written to the address. Therefore, the SMID which is the address where the matching data is written is output (SMID output), and the SMID is output.
The adding unit 23a adds the SMID to a predetermined position of the COM cell and sends it to the message processing unit 22 at the next stage.

If there is no match, a BOM cell having the same MID as the COM cell has not been input yet. In such a case, a malfunction has occurred, and a discard command CNC is output ( Disposal directive). When a discard command is issued, the cell discard unit 21c discards the corresponding cell.

(C) When an EOM cell arrives (see FIG. 13) When an EOM cell is input, the MID (1
Data DT indicating “0” and “in use” (= “0”) is created, and a search is made to determine whether data that matches the data is registered in the registration unit 21b-1. That is, starting from the head address (head SMID) of the registration unit 21b-1, a match search between the stored data and the data DT is sequentially performed, and a match is searched for (match search).

If there is a match, the MID of the BOM cell which is the first cell of the message to which the EOM cell belongs
(10 bits) has been written to the address. Therefore, the SMID which is the address where the matching data is written is output (SMID output), and the SMID is output.
The adding unit 23a adds the SMID to a predetermined position of the COM cell and sends it to the message processing unit 22 at the next stage.

Then, an all "1" pattern indicating unused is written in the above address, and the SMID is made unused (written). On the other hand, if no match exists, E
A BOM cell having the same MID as the OM has not been input yet. In such a case, since a malfunction has occurred, a discard command CNC is output (discard command). When a discard command is issued, the cell discard unit 21c discards the corresponding cell.

Message Processing Unit FIG. 14 is a block diagram of the message processing unit 22.
Stores the correspondence between the destination address DA and the routing information (virtual channel identifier: output VCI), and
Output destination address DA included in cell or SSM cell
DA-out VCI conversion unit for converting to CI, 22b is BOM
A routing information registration unit that stores the outgoing VCI (routing information) obtained by the conversion in accordance with the SMID added to the cell or the SSM cell, 22c is an outgoing VCI adding unit that adds the outgoing VCI to the cell, 22d is Outgoing MID assigning section for assigning a new MID to the output cell, 2
2e is an outgoing MID management unit that manages not to assign the same MID to each message in the service.
The SMID added to the cell or the SSM cell;
This is an output MID registration unit that stores a correspondence relationship with an MID assigned to an OM cell or an SSM cell.

Although not shown, the message processing unit 22
Has segment types ST and S from the input cell.
Means for separating MID, separated segment type ST
By decoding BOM cells, COM cells, EOM cells,
A segment type identification unit for identifying the SSM cell is included.

Routing control of message processing unit (a) When BOM cell or SSM cell is input When a BOM cell or SSM cell is input, the DA-V
The CI converter 22a obtains and outputs the output VCI from the destination addresses included in these cells. The routing information registration unit 22b stores the output VCI corresponding to the SMID added to the BOM cell or the SSM cell.

The outgoing VCI assigning section 22c outputs the outgoing VCI output from the DA-outgoing VCI converting section 22a to the BOM cell or the SOM.
Add to SM cell. In parallel with the above, the output MID management unit 22e obtains and outputs an unused MID. The outgoing MID registration unit 22f stores the MID corresponding to the SMID added to the BOM cell or SSM cell. The output MID assigning unit 22d outputs the MI output from the output MID management unit 22e.
D is added to the BOM cell or SSM cell and output.

(B) At the time of inputting a COM cell When a COM cell is input in such a state, an output VCI corresponding to the SMID added to the COM cell is obtained from the routing information registration unit 22b, and the MID is output. Determined by the registration unit 22f. Outgoing VCI providing unit 22c
Adds the obtained output VCI to the COM cell, and outputs
The ID assigning unit 22d assigns the obtained MID to the COM cell and outputs it.

(C) When an EOM cell is input When an EOM cell is input, the output information VCI corresponding to the SMID added to the EOM cell is obtained from the routing information registration section 22b, and the MID is output from the output MID registration section 2b.
Calculate from 2f. The output VCI adding unit 22c outputs the output V
CI is added to the EOM cell.
d assigns the obtained MID to the EOM cell and outputs it.

Thereafter, the routing information registration unit 22b
Deletes the output VCI corresponding to the corresponding SMID, and outputs
The ID registration unit 22f deletes the output MID corresponding to the SMID. As described above, the present invention has been described with reference to the embodiments. However, the present invention can be variously modified in accordance with the gist of the present invention described in the claims, and the present invention does not exclude these.

[0066]

As described above, according to the present invention, the number of in-server message identifiers SMI less than the message identifier MID
Since the MID and the routing information are configured to be stored in correspondence with D, the storage area in the server can be reduced. In particular, by setting the number of SMIDs to the number of messages that can be transmitted and received simultaneously, the storage area in the server can be reduced, and the SMID can be allocated to all messages in service.
Conversely, by limiting the number of SMIDs, the number of messages that can be transmitted and received simultaneously per subscriber can be limited.

Further, according to the present invention, EOM or SSM
When a cell to which is added is input, each temporarily stored correspondence is deleted, so that only the correspondence corresponding to the message that is always in service can be stored, and the memory capacity can be reduced. Also, according to the present invention, SMID is
Since it is configured to notify the message processing unit by adding it to an empty area of a cell or an SSM cell or an area not used in subsequent processing, the number of physical signal lines between the MID / SMID conversion unit and the message processing unit is reduced. Can be.

Further, according to the present invention, the use / non-use of each SMID is stored, and when a BOM cell or an SSM cell is input, an unused SMID is stored in the cell.
An SM that allocates an ID and makes it in use, and further releases the SMID by not using it after processing of an EMO cell is completed
MID registration / conversion for storing the correspondence between the MID and the SMID, and for converting the MID of the cell into the SMID using the correspondence between the MID and the SMID when the COM cell or the EOM cell is input, while storing the correspondence between the MID and the SMID. And a cell discarding unit that discards the cell when a cell having an MID not registered in the MID registration / conversion unit arrives, so that the SMID can be easily captured and released for each message, and belongs to the abnormal message. Cells can be easily detected and discarded.

According to the present invention, the use of each SMID /
Since the correspondence between the unused and the SMID and the MID is configured to be stored integrally, the conversion from the MID to the SMID and the management of the SMID can be physically realized by one mechanism.
At the same time, the storage capacity can be reduced.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of the present invention.

FIG. 2 is an overall configuration diagram of a communication system to which the connectionless communication system of the present invention is applied.

FIG. 3 is an explanatory diagram of an embodiment of an ATM format.

FIG. 4 is an explanatory diagram of a format of each layer.

FIG. 5 is a schematic configuration diagram of a server.

FIG. 6 is a configuration diagram of an SMID notifying unit.

FIG. 7 is a configuration diagram of a MID / SMID conversion unit.

FIG. 8 is a configuration diagram of an SMID management unit.

FIG. 9 is an explanatory diagram of storage data of each registration unit.

FIG. 10 is an explanatory diagram of stored data in a case where data is stored integrally.

FIG. 11 is an explanatory diagram of an operation when a BOM cell arrives.

FIG. 12 is a diagram illustrating an operation when a COM cell arrives.

FIG. 13 is an explanatory diagram of an operation when an EOM cell arrives.

FIG. 14 is a configuration diagram of a message processing unit.

FIG. 15 is a diagram showing a conventional example of a server connection method.

FIG. 16 is an explanatory diagram of a connectionless communication system.

FIG. 17 is a configuration diagram of a server.

FIG. 18 is an explanatory diagram of a server connection method.

[Description of Signs] 11. Connectionless information celling / restoring unit 12 Server 13 ATM network 21 MID / SMID conversion unit 22 Message processing unit 23 SMID notification unit

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Satoshi Kuroyagi 1015 Uedanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Prefecture Inside Fujitsu Limited (72) Inventor Haruo Mukai 1015 Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Fujitsu Limited ( 72) Inventor Ryuichi Takechi 1015 Kamiodanaka Nakahara-ku, Kawasaki City, Kanagawa Prefecture Inside Fujitsu Co., Ltd. (72) Inventor Tetsuya Nishi 1015 Kamiodanaka Nakahara-ku, Kawasaki City, Kanagawa Prefecture Inside Fujitsu Co., Ltd. -287536 (JP, A) JP-A-4-287541 (JP, A) JP-A-4-1388739 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04L 12/28 H04L 12/56 H04L 12/66

Claims (6)

(57) [Claims] A plurality of connectionless information cell conversion / restoration units (11) for performing bidirectional conversion between variable length connectionless information and connectionless cells as fixed length cells handled in an asynchronous transfer mode (ATM) network. ), A plurality of connectionless communication servers (12) provided in an exchange of the ATM network for analyzing the destination address in the connectionless cell, obtaining routing information of the connectionless cell and controlling the routing, and An ATM network (1) connecting between the servers (12) via the designated path
A) which divides connectionless information into cells and performs switching within the ATM network in connectionless cell units.
In a connectionless communication system in a TM network, a connectionless information celling / restoring unit (11) divides one message as variable-length connectionless information into a plurality of cells, and assigns the same message identifier to each cell. MID is added, and the first cell of the divided message has BOM as the segment type, the last cell has EOM, all the cells in the middle have COM, and one message is converted into only a single cell. Occasionally, an SSM is added, and the server (12) is set with a smaller number of in-server message identifiers SMID than the message identifier MID, and the BOM
Alternatively, when a cell to which an SSM is added is input, the message identifier MID of the cell is converted into an unused server message identifier SMID, the SMID is used, and the correspondence between the MID and the SMID is temporarily stored. In addition, the routing information in the ATM network is searched from the destination address in the connectionless information included in the cell to which the BOM or the SSM is added, and the routing information is temporarily stored in association with the SMID. When a cell to which EOM is added is input, the MID of the cell is converted into an SMID using the correspondence between the MID and the SMID, and the obtained routing information corresponding to the SMID is mapped to the correspondence between the SMID and the routing information. A in connectionless cell units via the path indicated by the routing information. Connectionless communication method in the ATM network to carry out the exchange of the M network. 2. The connectionless communication system in an ATM network according to claim 1, wherein, when a cell to which EOM or SSM is added is input, each of the temporarily stored correspondences is deleted. 3. The server (12), which converts a message identifier MID of a cell into an in-server message identifier SMID (MID / SMID conversion unit) (21) and performs a message processing such as a routing information search based on the SMID. A message processing unit (22) and a SMID for notifying the message processing unit of the SMID obtained by the MID / SMID conversion unit
The SMID notifying unit (23) includes a SMID notifying unit that adds the SMID to an empty area of a cell or an area not used in subsequent processing.
The adding unit (23a) and the SMID added by the SMID adding unit
SMID discriminating unit that discriminates and notifies the message processing unit
The ATM according to claim 2, comprising (23b).
Connectionless communication method in a network. 4. The MID / SMID conversion section (21) stores the use / non-use of each of the SMIDs and, when a cell to which a BOM or SSM is added is input, a message of the cell. Unused SMI for identifier MID
SM that allocates and outputs D and uses the SMID
The ID management unit (21a) stores the correspondence between MID and SMID, and
Or, when a cell to which EOM is added is input, the MID
And MID of the cell using the correspondence between
When a cell converted to D and to which EOM or SSM is added is input, the MID registration / conversion unit (21b) for deleting the correspondence between the MID and the SMID is not registered in the MID registration / conversion unit. 4. The connectionless communication system in an ATM network according to claim 3, further comprising a cell discarding unit (21c) for discarding a cell having an MID when the cell arrives. 5. Use / unuse of each SMID and SMID
5. A connectionless communication system in an ATM network according to claim 4, wherein the correspondence between the ID and the MID is stored integrally. 6. The number of SMIDs set is the number of messages that can be transmitted and received simultaneously per subscriber, and
2. The connectionless communication system in an ATM network according to claim 1, wherein the number of messages that can be transmitted and received simultaneously is limited by the number of IDs.