JP2002117013A - Multiprocessor communication system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve call processing capability and to efficiently manage resources so that a multiprocessor communication system, especially, an IMT-2000 system, etc., is adaptive to large-capacity traffic, calling at high frequency in a short period, a local traffic increase in a narrow area, etc. SOLUTION: A resource managing processor RMP which manages the whole resources allocate the respective resources to multiple call processors CLP respectively and each call processor performs call processing for acquiring and releasing a resource for individual calls by using resources allocated to itself, and also monitors the use state of the resources allocated to itself and makes a request to add or return an allocated resource through processing disconnected from the call processing operation according to the resource use state.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to a multiprocessor communication system. In recent years, with the rapid spread of mobile communication services, there has been a strong demand for the construction of a next-generation mobile communication system that enables high-speed and diverse mobile multimedia communication.

[0002] This next-generation mobile communication system is an ITU (International Telecommun- ication) for the purpose of multimedia, globalization, personalization and the like of mobile communication services.
Communication Union) standardized IMT-
2000 (International Mobile Telecommunication S
ystem-2000). In the IMT-2000 system, 600,000 BHCA (B
HCA: busy hour traffic), 3 million B at mobile switching gateway
HCA call processing capability is required.

The present invention provides a multi-processor communication system such as an IMT-2000 system switch so as to sufficiently cope with large-capacity traffic, short-term frequent calls, and local increase in traffic in a small area. The present invention relates to a resource capturing method and a multiprocessor communication device for improving call processing capability and improving resource management efficiency in a communication system.

[0004]

2. Description of the Related Art As one of the aims of an IMT-2000 system, the realization of multimedia communication has been proposed. The IMT-2000 system is required to perform a call connection process so as to enable various kinds of communication with various types of networks having different formats such as an existing public line network and a packet network.

FIG. 16 shows an outline of the configuration of an IMT-2000 communication system. A communication system 16-1 such as an exchange of an IMT-2000 system is an STM (Synchronous Tr
ansfer Mode (synchronous transfer mode) An STM switch (STM-SW) for performing exchange processing of the network 16-2 and an ATM (As
Asynchronous Transfer Mode Network 16
Switch (ATM-SW) that performs switching processing of the packet network 3 or the packet network 16-4, and an STM switch (STM-
SW) and an ATM switch (ATM-SW).
Switch (STM-SW) and ATM switch (ATM-
SW) for controlling a plurality of processors controlling the SW and SW).

An STM switch (STM-SW) switches a synchronous transfer mode main signal assigned to a predetermined time slot (TS) for each call to a destination time slot, and an ATM switch (ATM-SW) is asynchronous. The transfer mode ATM cell is identified by a connection identifier (VPI / V
Switching to the virtual path / virtual channel of the connection destination based on CI).

STM interface (STI1, 2)
Is the data transfer mode conversion, ie, the AT on the ATM network side.
Conversion between M cells and time slots on the STM network side (time slot-cell conversion) is performed. One STM interface (STI) accommodates 84 highways having 24 time slots, and the line capacity of one STM interface (STI) is 2016 (= 24 × 84) time slots.

[0008] In the IMT-2000 communication system, exchange processing is controlled by using a plurality of processors and dividing the control range of each processor into several exchange processing function groups (function distribution). Further, when the processing amount of the shared function group exceeds the processing capacity of one processor, the processing is shared among a plurality of processors (load distribution).

The processor types of the function group include a call processor (CLP) having a call processing function and a common line signaling processor (CS) having a common line signal processing function.
P), a maintenance operation processor (OMP) having a maintenance operation processing function, and a resource management processor (RMP) having a resource management function.

The processing sharing of each multiprocessor for one call is as follows. (I) Upon receiving a control signal for call connection control by making a call or the like, the common channel signaling processor (CSP) executes load distribution control on a plurality of call processors (CLPs) and executes the call such as the call. Select a call processor (CLP) to execute the processing. (Ii) The selected call processor (CLP) analyzes call connection control messages notified from the common channel signaling processor (CSP), translates numbers (dial numbers),
Executes call processing such as routing. (Iii) The resource management processor (RMP)
For call connection setting such as interface (STI),
It captures resources that are commonly used by multiple processors and manages their usage.

FIG. 17 shows an operation example of call connection processing by a conventional multiprocessor, and a call from the STM network is transmitted to an ATM.
The example which connects to a network is shown. The symbol (S *) added to the description of the following processing procedure corresponds to the symbol S * in the processing sequence shown in the figure.

First, a call request from the STM network is received by a common channel signaling processor (CSP) for executing common channel signal transmission / reception control (S1). The common channel signaling processor (CSP) distributes the call processing to one of the load processors (CLP2) for load distribution (SLP).
2) Send outgoing call information to the call processor (CLP2) (S3).

The call processor (CLP2) analyzes the received message of the call request and performs a protocol-level validity check process (S4), and performs digit translation, that is, digit analysis and translation of the caller number / callee number. And execute (S
5) The routing, that is, the connection route (path) is determined (S6), and the partner (opposite station) is specified (S7),
A resource capture processing request for the connection path is sent to the resource management processor (RMP) by inter-processor communication (S8).

The resource management processor (RMP) receives the resource capture request, performs a process of selecting a device (resource) to be allocated to path setting (S9), and refers to the resource allocation bitmap 17-1 to perform a resource capture process. Is executed (S10). Resource allocation bitmap 17-
1 is a time slot (TS) 0 of each highway (HW) 0 to 83 for each STM interface (STI).
23, the resource management processor (RMP) acquires an available empty time slot (resource) by referring to the empty / busy state, and notifies the call processor (CLP2) of the acquired resource. I do.

On the other hand, the call processor (CLP2)
A message editing process for transmitting to the partner station (ATM network) is executed (S11), and the common channel signaling processor (C
A call request is transmitted to SP2) (S12). The common channel signaling processor (CSP2) is connected to the partner station (ATM).
Message transmission processing to the network (S13,
S14).

A common channel signaling processor (CSP2)
Receives a response message to the message from the partner station (ATM network) (S15), and transmits the response message to the call processor (CLP2) (S1).
6). Upon receiving the response message, the call processor (CLP2) already has the resource management processor (RM
P) using resources captured by PTM-A
A path setting process between the TM networks is executed (S17).

Also, the call processor (CLP2)
The other station (AT) to the common channel signaling processor (CSP1)
(S18), and the common channel signaling processor (CSP1) transmits the response message to the STM network (S19). By executing the above-described call setting processing procedure, it becomes possible to communicate between the STM network side and the ATM network side.

[0018]

The function of each processor is designed in accordance with the characteristics and control allotment of the multiprocessor so that the above-described call setting processing procedure is executed, so that a path is set between networks of different formats. Communication becomes possible. However, as described above, the IMT-2000 system and the like are required to have the ability to process large-capacity traffic, high-speed outgoing calls, and the like at high speed.

In order to satisfy such requirements, it is necessary to greatly reduce the number of dynamic steps (the number of execution steps) in the call setting process and to significantly increase the processing capacity. The present invention improves a conventional resource capture process by inter-processor communication between a call processor and a resource management processor in a call setting process, and improves inter-processor communication between the call processor and the resource management processor during the call setting process. An object of the present invention is to reduce the load, improve the call processing capability, and improve the efficiency of resource management.

[0020]

A multiprocessor communication system according to the present invention comprises: (1) a plurality of call processors for executing call processing including acquisition and release of resources in response to a communication request, and resources for managing the entire resources. In a multiprocessor communication system including a management processor, the resource management processor assigns resources to the plurality of call processors in advance,
Each call processor performs call processing in the own processor using resources allocated to the own processor,
In addition, each call processor monitors the usage status of the resources allocated to the own processor, and, in accordance with the resource usage status, by a process separated from the call processing operation,
It requests addition or return of allocated resources.

Resources are allocated to a plurality of call processors in advance, and each call processor performs call processing in the own processor using the resources allocated to the own processor. Since the resource capture processing is performed in the own processor without sending a resource capture request directly to the resource management processor, the load of communication between processors during call processing can be reduced and the processing capacity can be improved.

(2) The resource management processor circulates a resource notification list storing the notification order and the allocation status of each resource between the call processors, and each call processor uses the resource notification list. This is to add or return the resources of the own processor.

By circulating through the resource notification list and adding or returning the resource allocation of each call processor, the resource management processor controls only the cycling of the resource notification list and delegates resource allocation processing to each call processor. Therefore, the processing load on the resource management processor can be reduced.

(3) The resource management processor, when detecting a device failure, stops sending the resource notification list and stops traversing the resource notification list during the device failure period.

(4) The resource management processor recognizes the newly installed call processor by inter-processor communication with the call processor, and incorporates the new call processor into the notification order of the resource notification list. The resource notification list is updated, and the updated resource notification list is circulated.

Also, (5) each of the call processors:
A request to add or return a resource to its own processor is sent to the resource management processor by inter-processor communication separated from the call processing operation, and the resource management processor adds or returns a resource to each call processor. In response to a request, resources are allocated to each call processor by inter-processor communication separated from the call processing operation.

[0027]

FIG. 1 shows the configuration of a multiprocessor communication system according to the present invention. As shown in FIG. 1, the communication system includes a hardware section and a software section. The hardware section is roughly divided into an STM exchange 1-1, an ATM exchange 1-2, and an STM interface (STI interface).
1-3).

STM interface (STI1-3)
Has a data transfer mode conversion function, and performs conversion between cells on the ATM exchange side and time slots on the STM exchange side. The line capacity of one STM interface (STI) is 2016 time slot (84HW × 24TS)
It is.

The STM exchange 1-1 is connected to a line device 1-11 having a line interface function with the STM network and a ST.
An M switch (STM-SW) and a signal termination device 1-12 are provided. The ATM switch 1-2 includes a line device 1-21 having a line interface function with an ATM network and an ATM.
A switch (ATM-SW) and a signal termination device 1-22 are provided.

The STM switch (STM-SW) switches the main signal of the synchronous transfer mode (STM) to the time slot of the connection destination, and the ATM switch (ATM-SW)
Uses the ATM cell as the connection identifier (VPI / VC
Switching to the virtual path (VP) / virtual channel (VC) of the connection destination based on I).

The software section includes a common channel signaling processor (CSP), a call processor (CLP), and a resource management processor (RMP). Each processor includes a hardware order control unit 1-3 for monitoring and controlling a hardware unit, and a device / processor management unit 1-4 for managing each device / process.

The common channel signaling processor (CSP) includes a protocol controller 1-5. Call processor (C
LP) includes a path (communication path) control unit 1-61 and a call control unit 1
-62, a resource management (individual) control unit 1-63, and an operation data (individual) management unit 1-64. The resource management processor (RMP) is a resource management (overall) control unit 1
-71 and an operation data (entire) management unit 1-72.

The protocol control unit 1-5 receives the call setup / release request message from the signal terminating devices 1-12 and 1-22, and checks the validity of the received message.
Further, upon receiving a message transmission request from the call control unit 1-62, the message is converted into an inter-station protocol format, and the message is transmitted to the signal terminating devices 1-12 and 1-22.

The call control unit 1-62 includes a protocol control unit 1
-5, a call setup / release request message is received, and the resource management (individual) control unit 1-63 is requested to capture a band (resource) required for communication, and the band (resource) is released at the end of communication. Make a request. Also, pass (call path)
A path setting / release request is made to the control unit 1-61.

The path (communication path) control section 1-61 receives a path setting / release request from the call control section 1-62, and receives an STM switch (STM-SW) or an ATM switch (ATM-S).
W), the path connection / release processing is executed, and the line device 1
-11 and 1-21 are transmitted.

The resource management (whole) control unit 1-71
Manages the assignment of resources to individual call processors (CLPs). According to the request for resource allocation / return from each call processor (CLP), the resource is provided / collected to each call processor (CLP).

The resource management (individual) control section 1-63 in each call processor (CLP) issues a resource allocation / return request to the resource management processor (RMP), and requests the resource management processor (RMP) for the resource allocated to its own processor. Manage resources of Call control unit 1 in own processor 1
When there is a request to acquire a band (resource) required for path setting from 62, the band (resource) acquisition process is executed within the range allocated to the own processor.

FIG. 2 shows the structure of data and operation data in the resource management (whole / individual) control unit. The resource capture idle / busy table in FIG.
P) is held as operational data, and for each STM interface (STI), a mounting flag indicating mounting / non-mounting of each highway, and an empty / busy state (allocation bitmap) of each time slot for each highway. Store.

The resource allocation table shown in FIG.
It is held as operation data in the resource management processor (RMP), and for each STM interface (STI), along with an implementation flag indicating implementation / non-implementation of each highway, the call processor (CLP) to which each highway is assigned Store the number.

The STI mounting / status table of FIG. 3C is held as operation data in the resource management processor (RMP) and the call processor (CLP),
For each TM interface (STI), an implementation flag indicating implementation / non-implementation of each highway, the number of the STM switch in which each highway is implemented, and the device state of each highway are stored.

The mounted call processor table of FIG. 4D is held as management data in the resource management processor (RMP), and stores the number of mounted call processors and the number of mounted call processors.

Further, the rotation flag data shown in FIG. 9E is stored as management data in the resource management processor (RMP), and is stored in each STM interface (STI).
The portable information is held every time. The resource notification list in FIG. 11F stores the order of notification to the call processors and the allocation status of each resource, and as general data common to each call processor (CLP), each call processor (CLP) and resource management Processor (RM
P). That is, each call processor (CLP)
And the resource management processor (RMP) carries around the resource notification list (F).

In the present invention, the management of all resources is performed by a resource management processor (RMP). Resources (highway time slots and the like) necessary for call connection are allocated to each implementation call processor (CLP) in advance. The process of capturing and releasing resources at the time of each call / end is delegated to each implementation call processor (CLP).

As an embodiment in which resources are appropriately allocated to each implementation call processor (CLP) in advance, [I]
Using the resource notification list (F) described above, resource data is circulated to each call processor (CLP), and a resource circulating method for allocating resources to each call processor (CLP); [II] A resource management processor for managing the entire resource There is a resource allocation method of calculating and allocating a resource allocation amount to be allocated to each implementation call processor (CLP) by (RMP).

[I] Resource Carrying Method First, the resource carrying method will be described. Under the main operation of the resource management processor (RMP), the resource data is sequentially transmitted to each of the mounted call processors (CLP) by the resource notification list (F). Using the circulated resource notification list (F), each mounted call processor (CLP) captures and secures necessary resources and returns unnecessary resources. Hereinafter, the processing procedure will be described.

[I-1] Procedure of Resource Carrying Process in Normal Operation FIG. 3 shows an example of operation of resource carrying process in normal operation. FIG. 4 shows a processing sequence of the resource management processor (RMP) in the operation example, and FIG. 5 shows a processing sequence of the call processor (CLP).

The resource management (overall) control unit 1-71 of the resource management processor (RMP) includes a device management unit 1-7.
A request for collection of the number of mounted STM switch devices is issued to the device No. 4 (S20). In addition, a request for collecting the number of the mounted STM interface (STI) devices between the STM and the ATM switch is made and acquired (S21).

Further, a resource management (whole) control unit 1-7
1 requests the processor management unit 1-73 to collect the number of mounted call processors and obtains it (S22). The processor management unit 1-73 and the device management unit 1-74
Is a management unit in the resource management processor (RMP) in the device / processor management unit 1-4 shown in FIG.

Next, a resource management (whole) control unit 1-7
1 generates a resource notification list (F) to be circulated by the implemented call processor (S23). Then, the carry flag data (E) is set to the ON (possible carry) state (S24).

Next, the resource notification list (F) is transmitted to the first call processor (CLP1) to be notified (S25). Upon receiving the resource notification list (F), the call processor (CLP1) updates the resource notification list (F) as appropriate, and updates the next call processor (CL).
Turn to P2).

The resource management (individual) control unit of the call processor (CLP1) includes a resource management processor (RM).
When the resource notification list (F) is received from the P) (or another call processor) (S26), the resource (time slot) usage status of the own call processor is transmitted from the operation data management unit 1-64 to the resource capture idle / busy table ( A)
(S27).

The resource management (individual) control unit 1-63
The usage rate of the resource (time slot) in the own processor is calculated (S28), and a comparison between the usage rate and the threshold is determined (S29). If the usage rate exceeds 80%, the available highway is captured with reference to the resource notification list (F), and the resource notification list (F) is updated to the captured state for the captured highway (S30). At the same time, it updates the resource capturing / busy table (A) as an individual resource newly allocated to its own call processor (S31).

When the usage rate is less than 50%, the resource is returned, the corresponding highway portion of the resource notification list (F) is updated to an empty state (S30), and the resource capture / occupancy table (A) is updated. Is updated to an unmounted state (S31).

The call processor (CLP1) notifies the resource notification list (F) of the next resource notification list (F) according to the notification order of the resource notification list (F).
P2) is specified, and the resource notification list (F) is transmitted to the call processor (CLP2) (S32).

Upon receiving the resource notification list (F), the call processor (CLP2) performs the same processing as that of the above-mentioned call processor (CLP1), and in accordance with the notification order instruction of the resource notification list (F), the call processor (CLP2). ) Is the last notification order, the resource notification list (F) is transmitted to the resource management processor (RMP) (S33), and the resource management (overall) control unit 1-7
1 receives the resource notification list (F).

The resource management (overall) control section 1-71
When the resource notification list (F) is received in a loop, it is confirmed that the resource notification list (F) is in the ON state (portable state) with reference to the rotation flag data (E) (S34). (F) is again transmitted to the call processor (CLP1) in the notification order (S35).

[I-2] Resource Carrying Process in Case of Device Failure During the execution of the resource allocation process in the above-described normal operation process procedure, the STM interface (STI) device failed for some reason and became unusable. In such a case, it is necessary to stop the resource allocation processing. In the following, a description will be given of the resource handover process when an STM interface (STI) device has failed and when the failure has been recovered.

FIG. 6 shows an operation example of the resource switching process when a device failure occurs and when a failure is recovered. FIG. 7 shows a processing sequence when a device failure occurs, and FIG. 8 shows a processing sequence when a device failure is recovered.

(I) Operation of Processor When Resource (STI Device) Fails The occurrence of a failure in the STM interface (STI) device is detected by the hardware unit (S36), and the occurrence of the failure is performed in the hardware order of the resource management processor (RMP). Control unit 1
-3 (S37), and the resource management (overall) control unit 1-71 of the resource management processor (RMP)
STM interface (STI) device failure information (number of failed STI device units)
-3 (S38).

The resource management processor (RMP) turns off the rotation flag data (E) of the faulty STM interface (STI) device (turns off) and temporarily stops the rotation of the resource notification list (F). Accordingly, after the resource notification list (F) is returned from the call processor (CLP), the circulation of the resource notification list (F) is stopped (S39).

Thereafter, the resource management processor (RM)
P) reads out the mounted call processor table (D) (S40) and notifies the mounted call processor (CLP) of the number of the failed STM interface (STI) device and the suspension of use (S41). Thereafter, until the failure of the STM interface (STI) device is restored, the resource transfer process for the corresponding STM interface (STI) device is not performed. For the STM interface (STI) device that is not in the failure, the resource rotation is continued.

When the call processor (CLP) receives from the resource management processor (RMP) a notification of the suspension of the use of the STM interface (STI) device, the call processor (CLP) sends the operation data management unit 1-63 from the resource management (individual) control unit 1-63. 64 is notified of resource use suspension (S42). The operation data management unit 1-64 reads the STI mounting / status table (C), and reads the STM interface (ST
I) Change the state of the device to the unusable state (S4)
3).

(Ii) Operation of Processor at Recovery of Resource (STI Device) Failure When the recovery of the failure of the STM interface (STI) device is detected by the hardware unit (S44), the recovery of the failure is performed by the resource management processor (RMP). Is notified to the hard order control unit 1-3 (S45). The failure recovery notification for each STM interface (STI) device number is transmitted to the resource management (entire) control unit 1-71 (S4).
6).

The resource management processor (RMP) reads out the mounted call processor table (D) (S4).
7) Notify the implemented call processor (CLP) of the restored STM interface (STI) device number and availability (S48).

Upon receiving the resource availability notification from the resource management processor (RMP), the call processor (CLP) notifies the resource management (individual) control unit 1-63 of the resource availability to the operation data management unit 1-64. (S49).

The operation data management unit 1-64 reads the STI mounting / status table (C) and changes the status of the STM interface (STI) device so that the capture process of the STM interface (STI) device can be restarted. A transition is made to a capture enabled state (S50). In addition, the resource management processor (RMP) sets the rotation flag data (E) to the ON state (rotation possible) (S51), and restarts the rotation processing of the resource notification list.

[I-3] Procedure for carrying around resources at the time of adding a call processor During operation of the communication system by performing resource allocation processing according to the above-described processing procedure, a call processor is newly incorporated and added (Plug- in). The procedure of the resource carryover process in such a case will be described below.

FIG. 9 shows an operation example of the resource carry-over processing when a call processor is added, and FIG. 10 shows a processing sequence of the operation example. Now, the resource management processor (RMP) → the call processor (CLP1) → the call processor (CLP2) → the resource management processor (R
(MP), while the resource notification list is in the process of rotating, the call processor (CLP) is newly added.
Assume that 3) is added and plugged in.

The processor management section 1-41 of the call processor (CLP3) has a resource management processor (RM).
P) is notified of the plug-in (S52), and the processor management unit 1-73 of the resource management processor (RMP) is notified.
Receives the notification of the plug-in from the call processor (CLP3) and the call processor number.

The processor management unit 1-73 of the resource management processor (RMP) having received the notification of the plug-in notifies the resource management (whole) control unit 1-71 of the plug-in of the call processor (CLP3) (S53). . The resource management (entire) control unit 1-71 updates the number of implemented call processors in the implemented call processor table (D), and adds the implemented call processor number (S54).

When the resource notification list (F) is returned from the call processor (CLP2) (S5
5), referring to the implementation call processor table (D),
It is detected that there is a change in the mounted call processor, and the notification order of the resource notification list (F) is edited again (S5).
6). Then, using the new resource notification list (F) in which the notification order has been re-edited, the resource transfer process for each call processor (CLP) is started (S57).

The call processor (CLP1), having received the resource notification list (F) from the resource management processor (RMP), captures the necessary highways and transmits the resource notification list (F) to the call processor (CLP3). (S58). Call Processor (CLP3)
Becomes idle after processing the plug-in,
The resource notification list (F) is waiting to be received.

When the call processor (CLP3) receives the resource notification list (F), the call processor (CLP3) captures an empty highway (HW) from the highway (HW) free / busy table of the list, and stores the empty highway (HW) in the resource capture / busy table (A). Assign (S
59). Then, the resource notification list (F) is transmitted to the next notification destination processor (S60).

[II] Resource Allocation Scheme Next, a resource allocation scheme according to the present invention will be described. Conventionally, the resource capture processing performed by the resource management processor (RMP) is replaced by a call processor (CL).
P) is similar to the above-described resource revolving method, but the allocation of resources to each call processor (CLP) is implemented by a resource management processor (RMP) that manages the entire resource. The resource allocation is calculated and distributed to each call processor (CLP).

Each call processor (C) to which resources have been assigned by the resource management processor (RMP)
LP) executes a time slot (band) acquisition process in the call connection process from the resources in the allocated range. FIG. 11 shows an operation example of the resource allocation processing, and FIGS. 12 and 13 show a processing sequence of the operation example.

[II-1] Operation of Resource Management Processor at System Restart The resource management (overall) control unit 1-71 of the resource management processor (RMP) executes the device management unit 1 at system restart.
In response to -74, a request for collecting and acquiring the number of mounted STM switches (S61), a request for collecting the number of mounted STM interfaces (STI) between the STM switch and the ATM switch.
An acquisition process (S62) is performed. Further, the processor management unit 1-73 executes a collection request / acquisition process of the number of mounted call processors (CLPs) (S63).

Next, a resource management (whole) control unit 1-7
1 calculates the number of resources to be allocated to each call processor (CLP) by the following formula (S64). Number of allocated resources = number of STIs × 2016 / number of implemented call processors The resource allocation information for each call processor (CLP) of the number of allocated resources according to the above formula is set in the resource allocation table (B) of the operation data management unit 1-72. (S65).

[II-2] Operation of Call Processor at System Resumption The resource management (individual) control unit 1-63 of the call processor (CLP) sends resource information (band information) to the resource management processor (RMP). (S66). The resource management processor (RMP) refers to the operation data 1-72, searches and edits the resource allocated to the call processor, transmits the allocated resource information to the call processor (CLP), and responds (S67). .

The call processor (CLP) sets the allocated resource information received from the resource management processor (RMP) in the resource capture availability table (A) of the operation data management unit 1-64 (S68).

[II-3] Resource allocation and release processing procedure Call processor (CLP) that receives a call request after performing the above resource allocation processing at the time of system restart.
Can perform resource capture (bandwidth capture) processing for the outgoing call only with the call processor (CLP).
This eliminates the need to send a resource capture request to the resource management processor (RMP) on a call-by-call basis (each call occurrence), reduces the number of dynamic steps in the call connection process, and reduces the processing capacity. Can be increased.

However, each call processor (CLP)
When the allocated resources are used up due to an increase in the traffic load, or when the allocated resources become unnecessary due to the reduction in the traffic load, it is necessary to newly perform appropriate resource reallocation processing. Becomes The operation of each processor for the resource reallocation procedure will be described below.

FIG. 14 shows an example of resource reallocation and release during system operation, and FIG. 15 shows a processing sequence thereof. The call processor (CLP1 / 2) manages resources (individually) when the system is idle while waiting for a processing request.
From the control unit 1-63 to the resource monitoring unit 1-65,
A monitoring task activation request is sent (S69).

The resource monitoring unit 1-65 performs resource usage monitoring processing in response to a monitoring task activation request (S7).
0). The task execution cycle of the resource usage monitoring process is
It can be every time slot allocation / release. The resource monitoring unit 1-65 includes an operation data management unit 1-65.
The resource allocation status is read from the 64 resource capture / occupancy table (A), the number of highway captures is determined (S71), and the resource usage rate is calculated by the following formula. Resource utilization = number of used time slots / number of allocated time slots.

The resource monitoring unit 1-65 compares the resource usage rate with the time slot capture threshold (S72). For example, if the resource usage rate is less than 60% or more than 80%, it is less than 60%. In the case of (1), a resource release request is transmitted to the resource management processor (RMP) in the case of 80% or more (S73).

The resource management (whole) control unit 1-71 of the resource management processor (RMP) responds to a resource capture / release request from each call processor (CLP).
A new resource allocation release process is performed (S7).
4). When a resource capture request is received from a certain call processor (CLP), the resource allocation table (B) of the operation data management unit 1-72 is searched for an unused highway, and the unused highway is newly allocated. Set as minutes, requesting call processor (CLP)
Notify. When a resource release request is received from a certain call processor (CLP), the corresponding highway is rewritten to an unused state with reference to the resource allocation table (B).

When the call processor (CLP) is notified of the newly allocated highway from the resource management processor (RMP), the call processor (CMP) designates the highway as an available highway and sets the resource capture idle / busy table (A) ) (S75).

[0087]

As described above, according to the present invention,
A resource is allocated in advance to each of the plurality of call processors, and each call processor performs call processing in the own processor using the resources allocated to the own processor. Since the resource capture processing is performed in its own processor without sending a capture request, the load of inter-processor communication during call processing is reduced,
Processing capacity can be improved.

In addition, resources can be efficiently managed by allocating the amount of resources to be allocated to each call processor according to the traffic load of each call processor by circulating the resource notification list or allocating resources by the resource management processor. Can be.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a multiprocessor communication system according to the present invention.

FIG. 2 is a diagram showing a configuration of data and operation data in a resource management (whole / individual) control unit.

FIG. 3 is a diagram illustrating an operation example of a resource carryover process in a normal operation.

FIG. 4 is a diagram showing a resource handling process sequence during a normal operation of a resource management processor (RMP).

FIG. 5 is a diagram showing a resource handling process sequence during a normal operation of a call processor (CLP).

FIG. 6 is a diagram illustrating an operation example of a resource switching process when a device failure occurs and when a failure is recovered.

FIG. 7 is a diagram showing a resource handover process sequence when a device failure occurs.

FIG. 8 is a diagram showing a resource handling process sequence at the time of device failure recovery.

FIG. 9 is a diagram illustrating an example of an operation of a resource roaming process when a call processor is added.

FIG. 10 is a diagram showing a resource carryover processing sequence when a call processor is added.

FIG. 11 is a diagram illustrating an operation example of a resource allocation process of a resource allocation method.

FIG. 12 is a diagram showing a resource allocation processing sequence of the resource allocation method.

FIG. 13 is a diagram showing a resource allocation processing sequence of the resource allocation method.

FIG. 14 is a diagram illustrating an operation example of resource reallocation and release of a resource allocation method.

FIG. 15 is a diagram showing a processing sequence of resource reallocation and release of the resource allocation method.

FIG. 16 is a diagram schematically illustrating a configuration of an IMT-2000 communication system.

FIG. 17 is a diagram showing an operation example of call connection processing by a conventional multiprocessor.

[Explanation of symbols]

1-1 STM exchange 1-2 ATM exchange 1-11, 1-21 Line equipment 1-12, 1-22 Signal terminator STI1, STI2, STI3 STM interface STM-SW STM switch ATM-SW ATM switch CSP Common line signal Method processor CLP call processor RMP resource management processor 1-3 hardware order controller 1-4 device / processor manager 1-5 protocol controller 1-61 path (communication path) controller 1-62 call controller 1-63 Resource management (individual) control unit 1-64 Operation data (individual) management unit 1-71 Resource management (whole) control unit 1-72 Operation data (whole) management unit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H04L 12/26 H04M 3/36 B H04M 3/00 H04Q 3/545 3/36 H04L 11/08 H04Q 3 / 545 (72) Inventor Kazue Kojima 3-9-118 Shin-Yokohama, Kohoku-ku, Yokohama-shi, Kanagawa Prefecture Inside Fujitsu Communication Systems Limited (72) Inventor Yukito Shibata 3-9-118 Shin-Yokohama, Kohoku-ku, Yokohama-shi, Kanagawa Prefecture F-term in Fujitsu Communication Systems Co., Ltd. (reference)

Claims (5)

[Claims] 1. A multiprocessor communication system including: a plurality of call processors that execute call processing including acquisition and release of resources in response to a communication request; and a resource management processor that manages all resources. Resources are allocated to the plurality of call processors in advance, each call processor performs call processing in the own processor using the resources allocated to the own processor, and each call processor allocates resources to the own processor. A multiprocessor communication system characterized by monitoring the usage status of the allocated resources and requesting addition or return of allocated resources according to the resource usage status by a process separated from the call processing operation. 2. The resource management processor circulates a resource notification list storing a notification order and an allocation state of each resource between call processors, and each call processor uses its resource notification list to execute its own processor. 2. The multiprocessor communication system according to claim 1, wherein resource addition or return is performed. 3. The resource management processor according to claim 1, wherein when detecting a device failure, the resource management processor stops sending the resource notification list and stops traversing the resource notification list during the device failure period. 3. The multiprocessor communication system according to 2. 4. The resource management processor recognizes a newly incorporated call processor by inter-processor communication with the call processor, and incorporates the new call processor into the notification order of the resource notification list to provide a resource notification list. 4. The updated resource notification list is circulated.
2. The multiprocessor communication system according to claim 1. 5. Each of the call processors sends a request to add or return a resource to its own processor to the resource management processor by inter-processor communication separated from a call processing operation. 2. A resource is allocated to each call processor by inter-processor communication separated from a call processing operation in response to a request to add or return a resource allocation of each call processor. Multiprocessor communication system.