JPH1168949A - Load distribution system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To most-efficiently distribute the load of each processor without causing the increase of a communication quantity between the processors concerning a load distributing system X at an exchanging network where the plural processors carry out processings by distributing the load among them. SOLUTION: The exchanging network 100 where the plural processors distribute the load to process is provided with a loading state monitoring means 101 monitoring the loading state of each processor, a distributing destination holding means 102 previously holding identification information of a prescribed another processor distributing the load added to a pertinent processor when an individual processor comes into a congested state, a executing load distributing processing distributing the load added to the pertinent processor to the processor of a load distributing destination held in the means 102 when a loading state monitoring means 101 detects a processor coming into a congested state.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a load distribution method in a switching network, and more particularly, to a load distribution method in a switching network in which a plurality of processing devices distribute and process a load.

An exchange is provided with a plurality of processing units each operating in a function distribution type or a load distribution type.
The switching network is provided with a plurality of exchanges each operating in a function distribution type or a load distribution type.

[0003] It is necessary to equalize the load on each processing device in these exchanges or each exchange in the exchange network as much as possible in order to improve the efficiency of the exchange or the exchange network.

[0004]

2. Description of the Related Art FIG. 10 is a diagram illustrating a conventional exchange. The exchange (1) shown in FIG. 10 includes a plurality of communication path devices (NW) (11) and a plurality of concentrators (LC) (1).
2), a plurality of digital terminals (DT) (13),
A call processor (CPR) (14) corresponding to each communication path device (NW) (11), and a set of main processor (MPR) (1)
5), a plurality of buses (16), a plurality of inter-channel adapters (CCAs) (17) and a set of maintenance terminals (SCWs).
S) and (18).

[0005] Each concentrator (LC) (12) accommodates a plurality of subscriber lines (2) and each digital terminal (DT).
(13) accommodates a plurality of trunk lines (3), each communication path device (NW) (11) accommodates a plurality of concentrators (LC) (12) or digital terminals (DT) (13),
A communication path connecting the subscriber line (2) and the trunk line (3) is formed.

Each of the call processing units (CPR) (14) disperses the load of a call occurring on the accommodation subscriber line (2) or the trunk line (3) and processes the call. The main processing unit (MPR) (15)
Performs the overall management of the exchange (1), and furthermore, the maintenance terminal (SCWS) (18)
And various commands for controlling the operation state, and outputs various messages indicating the operation state of the exchange (1).

[0007] Each channel adapter (CCA) (17)
Indicates that each of the call processing units (CPR) (14) and the main processing unit (M
PR) (15). An example (part 1) of a conventional load distribution method will be described for the exchange (1) shown in FIG.

For example, an arbitrary call processing device (CPR) (1
Assuming that the load of 4 ₁ ) exceeds the specified value and enters a congestion state, the call processing device (CPR) (14 ₁ ) takes charge of call processing under the management of the main processing device (MPR) (15). It is assumed that the call setting originating from each subscriber line (2) or trunk line (3) accommodated in the communication path device (NW) (11 ₁ ) is restricted.

In such a state, the communication path device (N
W) Call setting requests from all the subscriber lines (2) or trunk lines (3) accommodated in (11 ₁ ) are regulated. Eventually, when the load on the call processing device (CPR) (14 ₁ ) decreases below the specified value and the congestion state is resolved, the call processing device (CPR) (14 ₁ ) sets the main processing device (MPR) (1
Under the management of 5), the restriction on the call processing is released.

Next, an example (part 2) of a conventional load distribution method will be described for the exchange (1) shown in FIG. Each call processing device (CPR) (14) executes a call process occurring on each subscriber line (2) or trunk line (3) accommodated in the corresponding communication path device (NW) (11). .

Each call processing unit (CPR) (14)
While monitoring the congestion state of the own device, information indicating whether the own device is in the congestion state is transmitted to the main processing unit (MPR) (1).
5) The other congestion state of each call processing unit (CPR) (14) is notified to other call processing units (CPR) (14) via CPR, and notified from other call processing units (CPR) (14). Holding.

[0012] For example, when occurs the call setting request from any subscriber line (2 _1), call processor in charge of the call processing (CPR) (14 ₁₎ congestion condition of the subscriber line (2 ₁₎ If the other call processing device (CP
R) (14) is analyzed, and one of the other call processing devices (CPR) (14) that is not currently in a congestion condition is determined based on a predetermined standard [for example, the call processing device (CPR) (14 ₂ )]. ]
And requests the other selected call processing unit (CPR) (14 ₂ ) via the main processing unit (MPR) (15) for the call processing that has occurred on the subscriber line (2 ₁ ).

Requested call processor (CPR) (14 ₂ )
After confirming that the own device (14 ₂ ) is in a non-congested state, the device obtains information necessary for the call setting process, executes a required call setting process, and returns the execution result to the call processing device ( CP
R) (14 ₁ ) is notified.

Various information required for the call setting process is received from the requesting call processing device (CPR) (14 ₁ ), or the main processing device (MPR) (1) that commonly manages the information.
Receive from 5).

Next, an example (part 3) of the load distribution method will be described.
A description will be given of the exchange (1) shown in FIG.
Each call processing device (CPR) (14) sets up or releases a call occurring on each subscriber line (2) or trunk line (3) accommodated in the corresponding communication path device (NW) (11). The request is detected, but the call setup processing or the call release processing is executed by the slave processor (12) designated by the main processing unit (MPR) (15).

Each call processing device (CPR) (14)
The load status of the own device is monitored, the load status of the own device is notified to the main processing unit (MPR) (15), and the main processing unit (MP
R) (15) collectively manages the load status of each call processing device (CPR) (14).

[0017] For example, when a call setting request is originated from any subscriber line (2 _1), the call processing apparatus corresponding to the channel device that accommodates a subscriber line (2 ₁₎ (NW) (11 ₁₎ The (CPR) (14 ₁ ) detects the occurrence of the call setting request and notifies the main processing unit (MPR) (15) of the occurrence of the call setting request via the inter-channel adapter (CCA) (17).

The main processor (MPR) (15), which has received the notification of the occurrence of the call setting request, analyzes the load state of each of the sub-processors (12) managed by the processor and analyzes the load state of each of the slave processors (12). CPR) [for example, (14 ₂ )], and causes the selected call processing device (CPR) (14 ₂ ) to execute a call setting process generated from the subscriber line (2 ₁ ).

The information required for the call setting process is supplied from a central processing unit (MPR) (15) which is collectively managed, but the other call processing units (CPR) (14) are distributed separately. The held information is transferred to the main processing unit (MPR) (1
Transferring via 5) is also considered.

[0020]

As is apparent from the above description, in the conventional load distribution method (part 1),
The call processing device (CPR) (14 ₁ ) which has fallen into a congestion state
Even if the load is temporarily reduced due to the new call setting regulation and the congestion state is resolved, the call processing device (CPR) (14 ₁ ) will release the congestion state if the restriction is released unless the source of the load disappears. There was a problem that it occurred repeatedly and did not result in true load distribution.

In the conventional load distribution method (part 2), information indicating the congestion state of each of the call processing devices (CPR) (14) is always transmitted between the main processing device (MPR).
(15), and in the conventional load distribution method (part 3), each time a call setup and a call release occur, each call processing unit (CPR) (14) and the main processing unit (MPR) (15), it is necessary to send and receive various information related to call processing.
A) There is a problem in that the transfer information amount of (17) increases, and it becomes necessary to expand the information transfer capability of the inter-channel adapter (CCA) (17).

Further, in the conventional load distribution method (part 3), since the load of each call processing device (CPR) (14) is managed so as to be equal, all the call processing devices (CP) are controlled.
R) and (14) are simultaneously in a congestion state, and a state in which new call settings are simultaneously restricted occurs. In the case where the loads of all the call processing devices (CPR) (14) are equalized, Is a situation that was considered from the beginning, and if this is viewed as a problem, this method cannot be a solution to the problem.

An object of the present invention is to distribute the load of each processing device as efficiently as possible without increasing the amount of communication between the processing devices.

[0024]

FIG. 1 is a diagram illustrating the principle of the present invention. In the figure, reference numeral 100 denotes a switching network to which the present invention is applied.

101 is a switching network (100) according to the present invention.
This is a load state monitoring means provided in. Reference numeral 102 denotes distribution destination holding means provided in the switching network (100) according to the present invention.

103 is a switching network (100) according to the present invention.
This is a load distribution means provided in. Reference numeral 104 denotes a distribution destination determining means attached to the switching network (100) according to the present invention (claim 5).

In the switching network (100), a plurality of processing units distribute and process the load. Load state monitoring means (101)
Monitors the load status of each processing device. The distribution destination holding unit (102) holds, in advance, identification information of another processing device that distributes a load applied to the processing device when an individual processing device becomes congested.

The load distribution means (103) holds the load applied to the processing apparatus when the load state monitoring means (101) detects the processing apparatus in the congested state, in the distribution destination holding means (102). Load distribution processing for distributing the load to the processing device at the load distribution destination. [As above, the present invention (claim 1)] The distribution destination holding means (102) previously holds the identification information of a predetermined number of load distribution destination processing apparatuses, and stores the identification information of the load distribution destination processing apparatus at each time. It is considered to select from the identification information of the held load distribution destination processing devices according to a predetermined time lapse and a predetermined order. [Related to the Present Invention (Claim 2)] The load distribution means (103) records the number of distributed processing requests for each processing request added to each processing apparatus, and sets the number of distributed processing requests to a predetermined number. It is considered that the load distribution processing for the processing request that has reached the limit is regulated.
[Related to the Present Invention (Claim 3)] Also, the load distribution means (103) considers that only predetermined types of loads among the loads applied to the processing devices are to be subjected to the load distribution processing. You. [Related to the Invention (Claim 4)] The switching network (100) further comprises a load state monitoring means (101).
Collects the load status of each processing device monitored by the server, determines the load destination processing device of the load of each processing device based on the load status of each processing device and a predetermined procedure, and stores the destination storage device (102)
It is considered to add a distribution destination determining means (104) to be held in the storage device. [Related to the Present Invention (Claim 5)] Therefore, according to the present invention, the load status of a plurality of processing devices is analyzed in time series, and when each processing device becomes congested, the processing at the load distribution destination is performed. By selecting a device based on the above analysis results, it becomes possible to execute an effective load distribution process, it is also possible to prevent the traffic between the processing devices from being increased unnecessarily, and pay attention to the type of load, By executing the load distribution processing only for the load for which the load distribution processing is effective, the effective load distribution processing can be executed.
Further, if an appropriate procedure is set, it is not impossible to automatically determine the processing device to which the load is to be distributed from the load status of each processing device, and the processing capacity of the switching network is greatly improved.

[0029]

Embodiments of the present invention will be described below with reference to the drawings. 2 is a diagram illustrating an exchange according to an embodiment of the present invention, FIG. 3 is a diagram illustrating a primary table in FIG. 2, FIG. 4 is a diagram illustrating a secondary table in FIG. 2, and FIG. FIG. 6 is a diagram exemplifying the exponential stepping process in FIG. 2, FIG. 6 is a diagram exemplifying the load distribution process in FIG. 2, FIG. 7 is a diagram exemplifying the specific load distribution process in FIG. FIG. 9 is a diagram illustrating an exchange according to another embodiment of the present invention, and FIG. 9 is a diagram illustrating an exchange network according to an embodiment of the present invention (claim 5). The same reference numerals indicate the same objects throughout the drawings.

First, an embodiment of the present invention (claims 1 to 3) will be described with reference to FIGS. FIG. 2 shows a single exchange (1) as the exchange network (100) in FIG.

The exchange (1) has the same configuration as the conventional exchange (1) shown in FIG. 10, but each call processing unit (CPR) (14) has a primary table (91). , Secondary table (92), exponential step (9
3), utilization rate measurement unit (94) and load distribution unit (95)
Is provided.

The usage rate measuring section (94) plays the role of the load state monitoring means (101) in FIG. 1, and the primary table (91), the secondary table (92) and the index stepping section (93) Distribution destination holding means (102) in FIG.
The load distribution unit (95) further functions as a load distribution unit (103) in FIG.

The usage rate measuring unit (94) measures the usage rate (p _O ) of each of the own call processing units (CPR) (14), and uses the inter-channel adapter (CCA) (17) and the main processing unit (M).
PR) (15) through the maintenance terminal (SCWS) (1
8).

[0034] The maintenance person of the exchange (1) is in charge of each call processing equipment.
(CPR) (14) usage rate (p _OCollect) and constant
Each usage rate (p_O) Changes over time
For example, change every hour during the day, change every day during the week
, Change every month within the year) and analyze each call processor
(CPR) When (14) enters the congestion state, the load is reduced.
A call destination (CPR) (14) to be distributed is reserved.
And place them in the secondary table (92).
Put it.

FIG. 4 illustrates various secondary tables (92). 4 The (a), the secondary table in the case where the hourly day, time series the unit for a certain period of time (92 _A)
Is illustrated, and a specific call processing device (CPR) (1
In the case of 4), the call processing device (CPR) (14 _i1 ) is stored at the address corresponding to 0 o'clock as a load distribution request destination when the congestion state occurs at 0 o'clock every day. The call processing device (C
PR) (14 _i2 ) is stored at the address corresponding to 1 o'clock, and the call processing device (CPR) (14 _i3 ) is the address corresponding to 2 o'clock as a load distribution request destination when congestion occurs at 2 o'clock. Similarly, the call processing device (CPR) (14 _i24 ) is stored at the 23:00 corresponding address as a load distribution request destination when the congestion state occurs at the 23:00 time. An end code (STP) is stored in the address.

FIG. 4B shows a secondary table (92 _B ) when a fixed period is one week and a time series unit is a day of the week.
Is illustrated, and a specific call processing device (CPR) (1
The call processing device (CPR) (14 _j1 ) is stored at the address corresponding to Monday as a load balancing request destination when the congestion state occurs every Monday in 4), and the congestion state occurs on Tuesday. The call processing device (CPR) (14 _j2 ) is stored at the address corresponding to Tuesday as the load distribution request destination, and the call processing device (CPR) (14 _j2 ) is _transmitted as the load distribution request destination when the congestion state occurs on Wednesday. _j3 ) is stored in the address corresponding to Wednesday, and in the same manner, as a load distribution request destination in the case of congestion on Sunday,
The call processing unit (CPR) (14 _j7 ) is stored at the address corresponding to Sunday, and an end code (STP) is stored at the last address.
Is stored.

FIG. 4 (c) illustrates a secondary table (92 _C ) in which the fixed period is one year and the time series unit is monthly, and a specific call processing device (CPR) ( 14)
When the call processing device (CPR) (14 _k1 ) is stored in the address corresponding to January as a load distribution request destination when the system becomes congested in January every year, and the system becomes congested in February Processing Unit (CPR)
(14 _k2 ) is stored in the address corresponding to February, and as a load distribution request destination in the case of congestion in March,
The call processing device (CPR) (14 _k3 ) is stored at the address corresponding to March, and similarly, the call processing device (CPR) is used as a load distribution request destination when a congestion state occurs in December.
R) (14 _k12 ) is stored in the address corresponding to December, and the end code (STP) is stored in the last address.

The maintainer provides the various secondary tables (92) as described above and also provides a primary table (91) as shown in FIG. The primary table (91) shown in FIG. 3 includes a table mounting indication (TBL ₁ ), a command input / non-input identification call (CMD), and a data type (PT
N), exponent (IDX) and secondary table address (A
_TBL2 ).

The table mounting indication (TBL ₁ ) is information indicating whether or not the primary table (91) is mounted. The command input / non-input identifier (CMD) is information indicating whether or not a specific command has been input by the maintenance person.

The data type (PTN) is stored in the secondary table (9) provided in the call processing device (CPR) (14).
Indicates the type of 2) [see FIG. 4], for example, hourly secondary table (92 _A), depending on whether which of such day of week secondary table (92 _B) and monthly secondary table (92 _C) are provided , Data type (PTN = time), (PTN = day of the week), (PTN = month) and the like.

The index (IDX) is incremented by an index increment section (93), and a plurality of load balancing destination call processing units (CPR) (14) stored in each secondary table (92).
Specifying one that is currently selected from among, in hourly secondary table (92 _A), index (IDX) is incremented and 0 to 24 in synchronization with the respective times of the day, day of week in the secondary table (92 _B), index (IDX) is incremented and 0-7 in synchronization with a week day, in the month the secondary table (92 _C), index (IDX) is one year It advances from 0 to 12 in synchronization with each month.

The secondary table address (A _TBL2 ) is the secondary table (9) specified by the data type (PTN).
2) shows the storage area. Next, the process of stepping the index (IDX) by the index stepping unit (93) will be described with reference to FIGS.

For example, a call processing device (CPR) (14 ₁ )
Is provided with a secondary table (by time) (92 _A ), the exponential step-up unit (93) provides the call processing device (CP)
R) The change in time from 0 o'clock to 1 o'clock, 1 o'clock to 2 o'clock, and so on from 23 o'clock to 0 o'clock from the built-in clock mechanism of (14 ₁ ) as a factor for changing the index (IDX) Detect [FIG. 5, step S11].

In the initial state (ie, at 0 o'clock), it is assumed that the index is set to IDX (IDX = 0). In this state, when the clock mechanism outputs a change from 0:00 to 1:00,
Exponent stepping unit detects a change factor (93), the call processor (CPR) from the primary table (14) is held (91), the index (IDX = 0 o'clock) and the secondary table address (A _{TB L2)} Is extracted [Step S12].

Subsequently, the exponent step (93) generates an address (A _TBL2 + IDX + 1 = A _TBL2 +) generated from the extracted exponent (IDX = 0) and the secondary table address (A _TBL2 ).
1), the secondary table (by time) (92 _A ) is searched, and the identification information (14 _i2 ) of the stored load balancing destination call processing device (CPR) (14 _i2 ) corresponding to the index (IDX = 1) is retrieved. _i2 )
Detecting the [Step S13] When the detected load distribution destination call processor (CPR) identification information (14 _i2) of (14 _i2) is to ensure that not a termination code (STP) [Step S14], the primary table The index (IDX = 0) extracted from (91) is advanced to the index (IDX = IDX + 1 = 1) and stored in the index (IDX) storage area of the primary table (91) [Step S15].

Similarly, every time the clock mechanism outputs a change in time, the exponential step (93) sets the primary table (91).
Of the exponent (IDX) in the clock, the clock mechanism outputs the change from time 22:00 to 23:00, and the exponential step (9
When the clock mechanism outputs a change from time 23:00 to 0 o'clock with the index (IDX = 23) already set in the primary table (91) in the primary table (91), the index increment unit (9) detects the change factor.
3) executes the same process as described above (FIG. 5, steps S11 to S13), and generates an address generated from the index (IDX = 23) extracted from the primary table (91) and the secondary table address (A _TBL2 ). (A _TBL2 + IDX +
1 = A _TBL2 +24) and the hourly secondary table (92
_A ) is searched, and the stored end code (STP) is detected corresponding to the exponent (IDX = 24) [Step S13]. When it is confirmed that the search result is the end code (STP) [Step S14]. The index (IDX = 23) extracted from the primary table (91) is initialized (IDX = 0) and stored in the index (IDX) storage area of the primary table (91) [Step S16].

In the other day-by-day secondary table (92 _B ) or month-by-month secondary table (92 _C ), the index stepping unit (93) sets the index (IDX) to 0 to 7 in the same process.
[In the case of the secondary table (92 _B )], or 0 to 12
[In the case of the secondary table (92 _C )].

In the exchange (1) shown in FIG. 2, a specific information element (I) in a call processing request message transmitted / received via the subscriber line (2) and the trunk line (3).
E) [type of information element is omitted], a load distribution request number (n) for requesting load distribution to a predetermined load distribution destination call processing device (CPR) (14) is newly provided. (95) executes the load distribution request to the load distribution destination call processing device (CPR) (14) specified in the secondary table (92) because the local call processing device (CPR) (14) is in a congested state. Each time, 1 is added to the number of load distribution requests (n) in the call processing message, and when the number of load distribution requests (n) reaches a predetermined number of times (n _M ), further load distribution processing is performed. Regulating.

Next, a load distribution process according to an embodiment of the present invention (claims 1 to 3) will be described with reference to FIGS. 2 and 6, each call processor in the exchange (1) (CPR) (14 ), and that Hourly secondary table indicated (92 _A) are provided in FIGS. 4 (a) I do.

In this state, the communication path device (NW) (11)
_When a call setup request arrives at the exchange (1) from the subscriber line (2 ₁ ) accommodated in ₁ ), the call path device (NW) (11)
_When the call processing device (CPR) (14 ₁ ) corresponding to ₁ ) receives the arrived call setup request via the line concentrator (LC) (12 ₁ ) and the communication path device (NW) (11 ₁ ), [ 6, step S21], and activates the usage rate measuring unit (94 ₁ ).

The activated usage rate measuring unit (94 ₁ ) analyzes the current usage rate (p _O1 ) of the own call processing unit (CPR) (14 ₁ ) (step S22), and is determined for congestion determination. _Is compared with the threshold value (p _th ) [Step S2
3].

As a result of the comparison, if the current usage rate (p _O1 ) is less than the threshold value (p _th ), the local call processing device (CPR) (1
4 ₁ ) determines that the terminal is not in a congestion state, and notifies the call processing device (CPR) (14 ₁ ) of the determination result.

The call processing device (CPR) (14 ₁ ) reports the determination result that the device itself is not in the congestion state to the usage rate measuring section (9).
4), the local call processor (CPR) (14)
_{In 1} ), it is determined that the call processing can be executed, and the call processing in response to the received call setting request from the subscriber line (2 ₁ ) is executed [step S24].

As a result of the comparison, if the current usage rate (p _O1 ) is equal to or more than the threshold value (p _th ), the local call processing device (CPR)
(14 ₁ ) judges that it is in a congestion state, and notifies the result of the judgment to the call processing device (CPR) (14 ₁ ).

The call processing device (CPR) (14 ₁ ) measures the result of the determination that the device itself is in a congestion state by using a utilization rate measuring unit (9).
4), the call processing for the received call setting request from the subscriber line (2 ₁ ) is performed by the own call processing device (CP).
R) (14 ₁ ), it is determined that processing cannot be performed, and the load distribution unit (9
5. Start ₁ ).

The activated load distribution unit (95 ₁ ) analyzes the call setting request received by the call processing unit (CPR) (14 ₁ ) and determines the number of load distribution requests (n ) Is extracted [Step S25], and is compared with a predetermined number of times (n _M ) held in advance [Step S26].

As a result of the comparison, if the load distribution request number (n) extracted from the call setting request is less than the specified number (n _M ), the load distribution unit (95 ₁ ) determines that the call setting request has not yet been specified. It is determined that there is no problem even if the load distribution processing is requested to the predetermined other call processing device (CPR) (14) because the load distribution processing is not requested for the number of times (n _M ), and is extracted from the call setting request. Is added to the number (n) of load distribution requests (n = n)
After updating to +1), it is set in the predetermined information element of the call setting request [Step S27].

Next, the load distribution unit (95 ₁ ) refers to the primary table (91) provided in the call processing device (CPR) (14 ₁ ) (step S28), and stores the stored data type ( When it is determined that the PTN is time-dependent [Step S29], the call processing device (CPR) (14 ₁ )
There determines that already contains hourly secondary table (92 _A).

Next, the load distribution unit (95 ₁ ) extracts the secondary table address (A _TBL2 ) and the index (IDX) from the primary table (91), and extracts them from the call processing unit (CPR) (14).
₁₎ in the secondary table address (A _TBL2) for storing already hourly secondary table (92 _A), index (call processor corresponding to IDX) (CPR) (14 _iX) [where X is 1 to 24 either, for example, 14 _iX = 14 _2. ] Is extracted and selected as the call processing device (CPR) (14 ₂ ) of the load distribution request destination, and the call processing device (CP ₂ ) is extracted.
R) (14 ₁ ) is notified [step S2A].

When the call processing device (CPR) (14 ₁ ) receives the load distribution requesting call processing device (CPR) (14 ₂ ) notified from the load distribution unit (95 ₁ ), the call processing device (CPR) (14 ₁ ) receives the received subscriber line. The call processing in response to the call setting request from (2 ₁ ) is performed via the inter-channel adapter (CCA) (17 ₁ ) and the main processing unit (MPR) (15). ₂ ) [Step S2
F].

Call processing device (CPR) of load distribution request destination
(14 ₂ ) obtains information necessary for the call setting process after confirming that the own device is not in the congestion state, executes a required call setting process, and transmits the execution result to the call processing device (CPR) of the request source. )
Notify (14 ₁ ).

The call processing device (CPR) (14 ₁ ) controls the communication path device (NW) (11 ₁ ) based on the call processing result notified from the call processing device (CPR) (14 ₂ ) to join. The call setting request from the subscriber line (2 ₁ ) is satisfied.

In the above process [Step S26], the result of the load distribution unit (95 ₁ ) comparing the number of load distribution requests (n) in the received and analyzed call setting request with the specified number (n _M ). , The number of load distribution requests (n) is
_M ) or more, the load distribution unit (95 ₁ ) has been requested to perform the load distribution processing for the specified number of times (n _M ) for the call setting request. 1
4) the determination impossible request the load balancing process, and notifies the determination result to the call processor (CPR) (14 _1).

The call processing device (CPR) (14 ₁ ) receives the call setting request source subscriber line (2) based on the result of the determination from the load distribution unit (95 ₁ ) that the load distribution processing cannot be requested.
Inform ₁ ) that call setup is not possible.

As is apparent from the above description, according to the embodiments of the present invention (claims 1 to 3), the exchange (1)
The usage rate measuring unit (9) in each call processing unit (CPR) (14)
According to 4), the usage rate (p _O ) of the own apparatus is measured in time series and output via the main processing unit (MPR) (15). The measurement result of the usage rate (p _O ) is collected and analyzed, and when a congestion occurs, the call processing device (CPR) (14) of the load distribution request destination at each time is set in the secondary table (92). When an arbitrary call processing request occurs, the current usage rate (p _O ) of the local call processing device (CPR) (14)
Is compared with a predetermined threshold value (p _th ) to identify a congestion state. If a congestion state occurs, the secondary table (92)
The call processing device (CPR) (1
4), and the load distribution request destination call processor (CP) selected based on the statistical analysis.
R) (14) can be requested immediately for call processing.

The number of load distribution requests (n) is set for each call processing request, and 1 is added to each load distribution request. When the load distribution request becomes necessary, the load held by the call setting request is stored. The number of distributed requests (n) is compared with the specified number (n _M ),
When the number of load distribution requests (n) is equal to or greater than a specified number (n _M ), the load distribution requests are regulated to prevent invalid load distribution requests.

Next, an embodiment of the present invention (claim 4) will be described with reference to FIGS. When the exchange (1) accommodates mobile subscribers such as, for example, automobile telephones and mobile telephones, it becomes necessary to frequently execute processing specific to mobile communication, such as location registration processing or authentication processing. The occurrence of the congestion state caused by the mobile communication specific processing may greatly affect the efficiency of the exchange (1).

In FIGS. 2 and 7, the exchange (1)
Is a mobile communication specific process (for example, the above-described location registration process,
Authentication processing, etc.] are targeted for the load distribution processing. Each call processing unit (CPR) (14) in the exchange (1) has:
As before, it is assumed that Hourly secondary table indicated (92 _A) are provided in Figure 4 (a).

In this state, the communication path device (NW) (11)
_When a call setup request arrives at the exchange (1) from the subscriber line (2 ₁ ) accommodated in ₁ ), the call path device (NW) (11)
_When the call processing device (CPR) (14 ₁ ) corresponding to ₁ ) receives the arrived call setup request via the line concentrator (LC) (12 ₁ ) and the communication path device (NW) (11 ₁ ), [ FIG. 7, Step S31], and activates the load distribution unit (95 ₁ ).

The activated load distribution unit (95 ₁ ) analyzes the processing type received by the call processing unit (CPR) (14 ₁ ), and determines whether the received processing type is a processing type to be subjected to load distribution. It is determined whether or not it is [Step S3H].

[0071] the determination result, the reception process type is a call setup request, since the load distribution is other than mobile communication specific processing intended for the load distribution unit (95 ₁₎ of the target outside the load distribution processing type Is determined, and the result of the determination is determined by the call processing device (CP
R) (14 ₁ ) is notified.

The call processing device (CPR) (14 ₁ ) compares the result of the determination with the type not to be subjected to the load distribution processing with the load distribution unit (95).
₁ ), the call processing for the received call setting request from the subscriber line (2 ₁ ) is executed [step S3].
4].

Next, when a location registration request arrives at the exchange (1) from the subscriber line (2 ₁ ) accommodated in the channel device (NW) (11 ₁ ), the channel device (NW) (11 ₁ ). ) Is received by the call processing device (CPR) (14 ₁ ) via the line concentrator (LC) (12 ₁ ) and the communication path device (NW) (11 ₁ ) [step S31], and activates the load distribution unit (95 ₁ ).

The activated load distribution unit (95 ₁ ) analyzes the processing type received by the call processing unit (CPR) (14 ₁ ), and determines whether the received processing type is a processing type to be subjected to load distribution. It is determined whether or not it is [Step S3H].

As a result of the determination, since the reception processing type is a location registration request and a mobile communication specific processing to be load-balanced, the load distribution unit (95 ₁ ) The call processing device (CP
R) (14 ₁ ) is notified.

The call processing device (CPR) (14 ₁ ) sends the result of the determination as to the type of the load balancing process to the load balancing unit (9).
When notified from 5 ₁ ), the utilization rate measuring unit (94 ₁ ) is started for the purpose of determining whether or not the load distribution processing can be executed, as in the case of FIG. 6 described above.

The activated usage rate measuring unit (94 ₁ ) operates the own call processing unit (CPR) (1) in the same manner as described above (see FIG. 6).
The current usage rate (p _O1 ) of 4 ₁ ) is analyzed [Step S 3]
2], and compare with a threshold (p _th ) determined for congestion determination [step S33].

As a result of the comparison, if the current usage rate (p _O1 ) is less than the threshold value (p _th ), the own call processing device (CPR) (1
4 ₁ ) determines that the terminal is not in a congestion state, and notifies the call processing device (CPR) (14 ₁ ) of the determination result.

The call processing device (CPR) (14 ₁ ) determines the result of the determination that the own device is not in the congestion state by using the utilization rate measuring unit (9).
4), the local call processor (CPR) (14)
_{In 1} ), it is determined that the call processing can be executed, and the processing for the received location registration request from the subscriber line (2 ₁ ) is executed [step S34].

As a result of the comparison, if the current usage rate (p _O1 ) is equal to or greater than the threshold value (p _th ), the local call processing device (CPR)
(14 ₁ ) judges that it is in a congestion state, and notifies the result of the judgment to the call processing device (CPR) (14 ₁ ).

The call processing device (CPR) (14 ₁ ) determines the result of the determination that the own device is in the congestion state by using the utilization rate measuring unit (9).
4), the processing for the received location registration request from the subscriber line (2 ₁ ) is performed by the local call processor (CP).
R) (14 ₁ ), it is determined that processing cannot be performed, and the load distribution unit (9
5. Start ₁ ).

The activated load distribution unit (95 ₁ ) operates the call processing unit (CPR) (1
4 ₁ ) analyzes the received location registration request, extracts the number of load distribution requests (n) set in a specific information element [Step S35], and compares it with the prescribed number of times (n _M ) stored in advance. [Step S36].

As a result of the comparison, if the number of load distribution requests (n) extracted from the position registration request is less than the specified number (n _M ), the load distribution unit (95 ₁ ) determines that the position registration request has not yet been specified. It is determined that there is no problem even if the load distribution processing is not requested to the predetermined other call processing device (CPR) (14) because the load distribution processing is not requested only for the number of times (n _M ), and is extracted from the location registration request. 1 is added to the number of load distribution requests (n) of (n)
= N + 1), and then set it as a predetermined information element of the location registration request [Step S37].

Next, the load distribution unit (95 ₁ ) refers to the primary table (91) provided in the call processing unit (CPR) (14 ₁ ) [step S38], and stores the stored data type ( When it is determined that the PTN is time-dependent [Step S39], the call processing device (CPR) (14 ₁ )
There determines that already contains hourly secondary table (92 _A).

Next, the load distribution unit (95 ₁ ) extracts the secondary table address (A _TBL2 ) and the exponent (IDX) from the primary table (91), and extracts them from the call processing unit (CPR) (14).
Call processing device (CPR) (14 _iX ) corresponding to the index (IDX) of the hourly secondary table (92 _A ) stored in the secondary table address (A _TBL2 ) in ₁ ) [for example, 14 _iX =
14 ₂ to. ] Of extracting identification information (14 _2), and selected as the load distribution request destination of the call processor (CPR) (14 _2), and notifies the call processor (CPR) (14 ₁₎ [Step S3A].

When the call processing device (CPR) (14 ₁ ) receives the load distribution request destination call processing device (CPR) (14 ₂ ) notified from the load distribution unit (95 ₁ ), the received subscriber line (2 ₁₎ the processing of a position registration request from the inter-channel adapter (CCA) (17 ₁₎ and the main processor (MPR) call processor load distribution request destination through (15) (CPR) (14 ₂ ) [Step S3
F].

Call processing device (CPR) of load distribution request destination
(14 ₂ ) acquires the information necessary for the location registration process after confirming that the own device is not in the congestion state, executes the required location registration process, and transmits the execution result to the call processing device (CP) of the request source.
R) (14 ₁ ) is notified.

The call processing device (CPR) (14 ₁ ) controls the communication path device (NW) (11 ₁ ) based on the result of the location registration process notified from the call processing device (CPR) (14 ₂ ). The subscriber line (2 ₁ ) is notified of the completion of the location registration process.

In the above process, the load distribution unit (9
5 ₁ ) as a result of comparing the number of load distribution requests (n) in the received and analyzed call setting request with the specified number (n _M ) [Step S36], the number of load distribution requests (n) is determined to be the specified number (n).
_M ) or more, the load distribution unit (95 ₁ ) has been requested to perform the load distribution processing for the specified number of times (n _M ) for the position registration request, and the load distribution unit (95 ₁ ) further requests the other call processing device (CPR) ( 1
4) the determination impossible request the load balancing process, and notifies the determination result to the call processor (CPR) (14 _1).

[0090] The call processor (CPR) (14 _1), based on the determination result of the disabling request notified load distribution processing from the load distribution unit (95 _1), the location registration requesting subscriber line (2 ₁ ) Is notified that the location cannot be registered.

As is clear from the above description, according to the embodiment of the present invention (claim 4), the exchange (1) analyzes the processing type of the received processing request, and sets the processing to be the target of the load distribution processing. If it is determined that the type is a type (for example, a process specific to mobile communication), the same load distribution process as that of the present invention (claims 1 to 3) is executed. The load distribution processing is executed, and efficient load distribution processing can be executed.

Next, an embodiment of the present invention (claim 5) will be described.
This will be described with reference to FIGS. The switching network (100) shown in FIG. 9 includes a plurality of exchanges (1) and a set of host servers (5), and can communicate with each other via a common channel signaling network (4).

Each exchange (1) has the same configuration as that shown in FIG. 2, but in FIG. 9, a plurality of call processing units (CPR) (14) and a set of call processing Only a device (CPR) (14) and a common channel signaling device (CSE) (19) communicating via a common channel signaling network (4) are shown, and other components and each call processing are shown. A primary table (91), a secondary table (92), an exponential step (9) provided in the device (CPR) (14).
3), utilization rate measurement unit (94) and load distribution unit (95)
Has been omitted.

The host server (5) includes an information processing section (51) and a common channel signaling device (CSE) (52). The information processing section (51) further includes a usage rate collecting section (96). ), A distribution destination determining unit (97), and a secondary table updating unit (98).

The usage rate measuring unit (94) in each call processing unit (CPR) (14) provided in each exchange (1) is shown in FIG.
Each own call processing unit (CPR) (14)
Utilization of (p _O) periodically measured, and outputs the unexpected shown maintenance terminal (SCWS) (18), and holds the measured usage rate of (p _O).

On the other hand, the usage rate collection unit (96) in the information processing unit (51) periodically accesses each exchange (1) via the common channel signaling network (4), and Each usage rate (p) measured and held by each usage rate measuring unit (94) in each call processing device (CPR) (14) provided in
_O ) Collect.

The distribution destination determining unit (97) statistically processes the usage rate (p _O ) of each call processing device (CPR) (14) collected by the usage rate collection unit (96), and processes each call processing device. (CP
R) (14) to analyze the time-series fluctuations of the usage rate (p _O ), and call processing equipment (CPR) (14) in each exchange (1).
Not only in each case but also in some cases, the call processing device (CPR) (14) in each exchange (1), the call processing device (CP) of the request destination of the load balancing process for executing the load balancing process.
R) and (14) are analyzed in time series, and each call processing device (CP
R) The storage order of the secondary table (92) in (14) is
Determined based on a predetermined procedure.

The call processing device (CP
R) (14) as an example of a procedure for determining the load distribution request destination call processing device (CP) of the call processing device (CPR) (14) having the maximum usage rate (p _O ) in the same time series section.
R) (14), the call processing device (CPR) (14) having the minimum usage rate (p _O ) is determined, and the maximum usage rate (p _O ) among the remaining call processing devices (CPR) (14) is determined. The minimum usage rate (p _O ) of the call processing device (CPR) (14) of the call processing device (CPR) (14) having the
The call processing device (CPR) (14) having the load distribution request is determined in the same manner.
(14) and the call processing device (CPR) of the load balancing request destination
(14) are sequentially determined.

Also, the call processing apparatus (CP
R) (14) and the call processing device (CP
In the case where R) and (14) are determined between different exchanges (1), for example, it is applied to a specific processing type (the above-described location registration processing in an exchange network for mobile communication, or authentication processing, etc.) , A reference is provided in advance.

The secondary table updating unit (98), based on the determination result of the distribution destination determining unit (97), stores the secondary table in each call processing unit (CPR) (14) provided in each exchange (1). The updated contents of the table (92) are transmitted to the common line signaling device (CS).
E) Transfer to each exchange (1) via (52) and common line signaling network (4), and common line signaling device (CSE) (19) and main processing unit (MPR) in each exchange (1) (1
The contents of the secondary table (92) in each call processing device (CPR) (14) are updated via 5).

The load distribution processing between the exchanges (1) as described above is performed on individual exchanges (1), for example, processing specific to the mobile communication exchange network (the above-described location registration processing or authentication processing, etc.). It is effectively used for processes that are relatively loosely coupled.

As is apparent from the above description, according to the embodiment of the present invention (claim 5), instead of the maintenance person executing the configuration of the secondary table (92) one by one, an appropriate procedure is determined and distributed. If the secondary table (92) can be automatically set and updated if it can be set in the destination determination section (97), not only can the labor of the maintenance person be reduced,
It is also possible to eliminate artificial errors.

FIGS. 2 to 9 are only one embodiment of the present invention, and for example, a call processing device (CPR) (14)
Is the secondary table (92) is the hourly secondary table (9)
2 _A) without to be limited to those which store the day of week secondary table (92 _B), monthly secondary table (92 _C)
Although many other deformations are considered, such as storing, the effect of the present invention does not change in any case. In addition, exchange (1)
Is not limited to storing only the hourly secondary table (92 _A ), but the day of the week secondary table (9A).
2 _B) and monthly secondary table (92 _C) also features, by specifying the data type of the one kind of the primary table (91) currently used therein (PTN), etc. to be used by appropriately selecting , But many other variants are considered,
In any case, the effect of the present invention does not change. Further, the primary table (91), the secondary table (92) and the exponential stepping unit (93) are not limited to those stored in each call processing unit (CPR) (14), and are shown in FIG. 8, for example. As described above, the call processing device (CPR) (14) that stores the information in common with the main processing device (MPR) (15) and performs a load distribution request
Each time, various modifications are considered, such as referring to the primary table (91) and the secondary table (92) in the main processor (MPR) (15). The effect remains the same. It goes without saying that the switching network (100) to which the present invention is applied is not limited to the illustrated exchanges or switching networks.

[0104]

As described above, according to the present invention, in the switching network, the load status of a plurality of processing devices is analyzed in time series,
When each processing device is in a congestion state, by selecting a processing device of a load distribution destination based on the above-described analysis result,
Effective load distribution processing can be executed, and the traffic between the processing devices can be prevented from being increased unnecessarily. By executing the distributed processing, effective load distribution processing can be executed. If an appropriate procedure is set, it is not possible to automatically determine the processing apparatus to which the load is to be distributed based on the load status of each processing apparatus. It is no longer possible, and the processing capacity of the switching network is greatly improved.

[Brief description of the drawings]

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

FIG. 2 shows an exchange according to an embodiment of the present invention.

FIG. 3 is a primary table in FIG. 2;

FIG. 4 is a secondary table in FIG. 2;

5 is an exponential step processing in FIG.

FIG. 6 shows the load distribution processing in FIG.

FIG. 7 is a specific load distribution process in FIG.

FIG. 8 shows an exchange according to another embodiment of the present invention.

FIG. 9 is a switching network according to an embodiment of the present invention (claim 5);

FIG. 10 shows a conventional exchange.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Switch 2 Subscriber line 3 Trunk line 4 Common line signal network 5 Host server 11 Communication line device (NW) 12 Concentrator (LC) 13 Digital terminal (DT) 14 Call processor (CPR) 15 Main processor (MPR) Reference Signs List 16 bus 17 inter-channel adapter (CCA) 18 maintenance terminal device (SCWS) 19, 52 common channel signaling device (CSE) 51 information processing unit 91 primary table 92 secondary table 93 index stepping unit 94 utilization measuring unit 95 load distribution Unit 96 usage rate collecting unit 97 distribution destination determining unit 98 secondary table updating unit 100 switching network 101 load state monitoring unit 102 distribution destination holding unit 103 load distribution unit 104 distribution destination determining unit

──────────────────────────────────────────────────の Continued on front page (51) Int.Cl. ⁶ Identification code FI H04Q 3/545

Claims (5)

[Claims] In a switching network in which a plurality of processing devices distribute and process a load, a load status monitoring means for monitoring a load status of each of the processing devices, and when each of the processing devices becomes congested, A distribution destination holding unit that holds in advance identification information of another processing device that distributes a load applied to the processing device; and a load destination monitoring unit that is added to the processing device when the load state monitoring unit detects a processing device in a congested state. A load distribution means for executing a load distribution process for distributing a load to a processing device of a load distribution destination held by the distribution destination holding means. 2. The load distribution destination processing device according to claim 1, wherein the distribution destination holding unit retains identification information of a predetermined number of load distribution destination processing devices in advance, and stores the identification information of the load distribution destination processing device at each time. 2. A method according to claim 1, wherein the information is selected in accordance with a predetermined time and a predetermined order from the identification information.
The load balancing method described. 3. The load distribution means records the number of distributed processing requests for each processing request added to each of the processing devices, and stores the number of distributed processing requests when the number of distributed processing requests reaches a predetermined number. The load distribution method according to claim 1, wherein the load distribution processing is regulated. 4. The load according to claim 1, wherein the load distribution unit targets only a predetermined type of load among loads applied to each of the processing devices, as a load distribution processing target. Distributed method. 5. The switching network collects the load status of each of the processing devices monitored by the load status monitoring means, and based on the load status of each of the processing devices and a predetermined procedure, determines the load of each of the processing devices. 2. The load distribution method according to claim 1, further comprising a distribution destination determination unit that determines a distribution destination processing device and causes the distribution destination holding unit to hold the distribution destination processing device.