JP5944336B2 - Communication system, management server, and control program - Google Patents

Description

  Embodiments described herein relate generally to a communication system, a management server, and a control program.

  Currently, there is a movement to replace an existing public network with an IP network by a next-generation IP network technology called NGN (Next Generation Network). Telecommunications carriers are promoting IP and optical communications, and the number of IP telephone subscribers is increasing year by year. In addition, as represented by cloud technology, there has been an increasing demand for devices to be freely distributed on an IP network.

  On the other hand, when the number of devices increases in a system in which devices are distributed on an IP network, a large number of network devices simultaneously access a control device such as a call control server having a call connection processing function as an example. It is assumed that In such a case, a load that exceeds the tolerance of the apparatus is caused, and a problem occurs in that the processing performance is reduced or the apparatus is down. Therefore, in a system that is distributed and arranged on an IP network, load distribution when the number of network devices increases becomes a major issue.

JP 2009-141826 A

  As a conventional load distribution method, DNS (Domain Name System) that distributes connection destinations by assigning multiple IPs to the same domain name and using them alternately, and distributing the load by a device on the network side that receives an access request There is a method using a load balancer. Many of the load distribution methods take the form of distribution after receiving an access request. However, passive load distribution cannot prevent access load concentration on a device that executes load distribution.

  Therefore, load balancing methods that switch the access destination so as to actively distribute the load from the network device on the connection side can be considered, but these methods perform load balancing based on instantaneous load status, so the load balancing destination It is considered that it is not possible to cope with a sudden increase in load in the apparatus.

  Patent Document 1 discloses a load distribution method for determining an area where the number of terminals increases based on a statistical access frequency in mobile communication and reducing the load of location registration by changing the configuration on the core network side. Has been proposed.

  However, this method only responds by predicting the location registration frequency of the area from the statistical information and increasing resources on the core network side, and there is a possibility of selecting an inappropriate device as a load distribution destination. That is, when a device whose load tends to increase is selected as a load distribution destination, load distribution processing frequently occurs, and it may be difficult to guarantee stable communication.

Communication system embodiment includes a plurality of network devices, a plurality of control devices, and a management server, a.
Controller respectively controls communication Netw network devices assigned thereto via a communication network.
In parallel with this, the management server manages a plurality of control devices via the communication network, and each network is based on the current load status accompanying the communication control of the control device and the temporal transition of the load in each network device. It estimates the geographical environment in which the device is installed, based on the estimated condition of future burden of communication control of the control device which is predicted by the geographical environment, re Shi pairs to each control unit Netw network device A load distribution instruction for assigning is performed.

FIG. 1 is a schematic configuration block diagram of a communication system according to an embodiment. FIG. 2 is an explanatory diagram of a schematic configuration of each unit constituting the communication system. FIG. 3 is an explanatory diagram of an example of an operation when collecting call information of the network device. FIG. 4 is an explanatory diagram of an example of a call information storage table in which collected call information is stored. FIG. 5 is an explanatory diagram of an example of a predicted load status value operation when load distribution is performed using throughput as an index. FIG. 6 is an explanatory diagram of past statistical information. FIG. 7 is an explanatory diagram of an example in the case of determining load distribution based on the load status of the control device using throughput as an index. FIG. 8 is an explanatory diagram of an example of a connection destination changing operation in load distribution. FIG. 9 is an explanatory diagram illustrating an example of a case where the control device that has received the load distribution instruction selects a network device that changes the connection destination. FIG. 10 is an explanatory diagram of an example of a communication sequence related to load distribution. FIG. 11 is an explanatory diagram of an example in which the geographical environment of the network device is estimated based on the call information. FIG. 12 is an explanatory diagram of an example in the case of determining load distribution based on the load status of the control device using the throughput and the geographical environment as indices. FIG. 13 is an explanatory diagram of an example of a connection destination changing operation in load distribution according to the fourth embodiment. FIG. 14 is an explanatory diagram of an example of a communication sequence related to load distribution in the fourth embodiment. FIG. 15 is an explanatory diagram of an example in the case of determining load distribution based on the load status of the control device using the throughput, the temporal trend of the throughput, and the geographical environment as an index.

Next, embodiments will be described with reference to the drawings.
[1] First Embodiment FIG. 1 is a schematic configuration block diagram of a communication system according to an embodiment.
The communication system 10 includes a plurality (two in FIG. 1) of network devices 11-1 and 11-2 configured as an information processing apparatus such as a personal computer, and a communication network 12 configured as an IP network and a public communication network. A plurality of (two in FIG. 1) control devices 13-1, which are connected to the network devices 11-1 and 11-2 via the communication network 12 and perform the requested call connection processing to establish communication. 13-2 and a management server 14 that collects call information accumulated by the control devices 13-1 and 13-2, manages load conditions of the control devices 13-1 and 13-2, and executes load distribution processing. It is equipped with.

In the above configuration, the network devices 11-1 and 11-2, the control devices 13-1 and 13-1 and the management server 14 are connected to the communication network 12. The control devices 13-1 and 13-2 manage the call information of the network devices 11-1 and 11-2, manage the network devices 11-1 and 11-2, and specify the connection destination at the time of call connection. Control.
As described above, the communication system 10 according to the present embodiment constructs a cloud communication system in which a plurality of control devices 13-1 and 13-2 and the management server 14 cooperate to perform communication control as a whole. Yes.

FIG. 2 is an explanatory diagram of a schematic configuration of each unit constituting the communication system.
The network device 11-1 and the network device 11-1 have the same configuration, and each includes a call connection processing unit 21 that processes call connection via the communication network 12.

  Further, the control device 13-1 and the control device 13-2 have the same configuration, and the call control unit 31 that controls the call connection processing of the network device and the call information CI related to the network device that is in communication, respectively. And a connection for switching the connection destination of the network devices 11-1 and 11-2 based on the load distribution instruction of the management server 14 and the call information management unit 32 for notifying the call information in response to a request of the management server 14 A destination control unit 33 and a call information storage table 34 in which call information CI is accumulated are provided.

In addition, the management server 14 collects and stores the call information CI accumulated by the control devices 13-1 and 13-2, and the network devices 11-1 and 11 based on the collected call information CI. -2 is installed to estimate the geographical environment where the network environment 11 is installed, and the load status of the network devices 11-1 and 11-2 is statistically processed in time to grasp the temporal trend of the load status. The time statistics calculation unit 43, the load distribution instruction unit 44 that instructs load distribution based on the collected information, the geographical environment information GI for each network device estimated by the geographical environment estimation unit 42, and the time statistics calculation unit 43 And a database 45 for accumulating the calculated temporal statistical information SI.
Here, the geographical environment information GI and the statistical information SI accumulated in the database 45 are used for load distribution determination in the load distribution instruction unit 44.

FIG. 3 is an explanatory diagram of an example of an operation when collecting call information of the network device.
In the following description, the network device 11-1 performs voice communication via the control device 13-1, and the network device 11-2 performs data communication via the control device 13-2. Shall.
The control devices 13-1 and 13-2 collect various information related to communication for each communication session, and store the call information CI. The collected call information CI is managed as statistical information for each of the network devices 11-1 and 11-2 in association with the network device ID for specifying the corresponding network devices 11-1 and 11-2.
The management server 14 periodically transmits a call information acquisition request to the control devices 13-1 and 13-2, collects the call information CI accumulated by the control devices 13-1 and 13-2, and integrates them. to manage. The collected call information CI is managed for each of the control devices 13-1 and 13-2, stored in the database 45 as the geographic environment information GI by the geographic environment estimation unit 42, and temporal statistical information SI by the temporal statistics calculation unit 43. Is stored in the database 45.

FIG. 4 is an explanatory diagram of an example of a call information storage table in which collected call information is stored.
When the communication session between the network devices 11-1 and 11-2 and the control devices 13-1 and 13-2 is disconnected, various types of communication information are stored together with the call information ID for specifying the call information CI. Stored in the table 34.

The call information storage table 34 is roughly divided to store network device ID data 34A for storing network device IDs for specifying the network devices 11-1 and 11-2, and call information ID for specifying the call information CI. Call information ID data 34B, date data 34C for storing a communication occurrence date, communication time data 34D for storing a communication start time and a communication end time, and a call storing a call type (voice call or data communication) Type data 34E, throughput data 34F storing average communication throughput, priority data 34G storing communication priority, and the like are stored.
In the above configuration, the call information ID is uniquely assigned to each communication session in the control devices 13-1 and 13-2.

As the communication occurrence date, the date when the communication is started is stored. When the session is established, the communication start time is stored in the communication time data 34D, and when the session is disconnected, the end time is stored in the communication time data 34D.
Further, the call type such as voice communication and data communication is determined based on the call control message transmitted from the network device at the time of call connection, and stored in the call type data 34E.

Further, the average throughput of the communication period is calculated when the session is disconnected and stored in the throughput data 34F.
In IP packet communication, in particular, when a priority parameter such as TOS (Type Of Service) is set, priority information is also held in the priority data 34G.
In addition, what was mentioned above is an example of call information CI, and does not limit the information which can be collected with the control apparatuses 13-1, 13-2 and the management server 14.

FIG. 5 is an explanatory diagram of an example of a predicted load status value operation when load distribution is performed using throughput as an index.
FIG. 6 is an explanatory diagram of past statistical information.
The management server 14 holds past temporal statistical information related to throughput for each of the network devices 11-1 and 11-2.
In the example of FIG. 5, the past load situation is held as a time averaged throughput every hour. Therefore, the next 12:00 load situation is predicted to determine the necessity of load distribution.

The management server 14 searches for data similar to the current load status based on the past statistical information (history) shown in FIG.
As past statistical information to be compared for determining the similarity, as shown in FIG. 6, the same time zone of the previous day, the same time zone on the same day of the week before, and the same time zone of one year ago are candidates. The management server 14 compares the past information with each other, and determines the degree of similarity for the load status for a certain period. The determination of the similarity is calculated, for example, by selecting the one with the smallest sum of differences between the current day data and the past same time data.

  In FIG. 5, the similarity is determined from the throughput of the past 4 hours (the estimated time is 12:00, so 8:00, 9:00, 10:00, and 11:00). Is selected as past reference data RE, that is, data of the same day that is most similar to the data of the current day.

  More specifically, as shown in FIG. 5, “31.0 Mbps”, which is the same day data having the highest similarity to the data of the current day, is used as the past reference data RE, and the load status estimate of the network device is Then, 36.0 Mbps obtained by adding a preset error threshold of 5 Mbps to the past reference data RE is calculated as the estimated data ES as the load state expected value.

In this case, the error threshold value may be given as an initial value by the system designer based on statistical information in advance, or a variance value may be calculated based on throughput data for a certain period and used as an index of the error threshold value. Good.
In addition, a threshold is set for the similarity determination, and when the sum of the differences between the current day data and the past same time data exceeds the threshold for similarity determination, a preset load situation reference value is set as the load situation prediction load. It is also possible to use it as an estimated load status.

Furthermore, the load condition reference value may be given in advance as an initial value by the system designer based on statistical information, or may be automatically calculated as an average throughput of all past statistical information.
In the example of FIG. 5, an example is shown in which the estimated data ES is calculated as the estimated load status based on the average throughput history. However, the load status is estimated based on the call connection frequency or the CPU processing load of the control device. It can also be used as an index of value.

FIG. 7 is an explanatory diagram of an example in the case of determining load distribution based on the load status of the control device using throughput as an index.
The management server 14 calculates the expected load status of each of the network devices 11-1 and 11-2 in the load distribution processing operation. Then, based on the calculation result of the expected load status, the control device (13-1, 13-2) in which the sum of the expected load status of the connected network devices exceeds the load distribution target threshold is distributed to the other control devices. It is assumed that the load balancing source control device is a target to be executed.

In addition, control devices (13-1, 13-2) whose estimated total load status of the network devices to be connected is less than the stable operation threshold are set as candidates for load balancing destination control devices.
In the case of the example in FIG. 7, the load distribution target threshold is 200 Mbps, and the stable operation threshold is 50 Mbps.
More specifically, the management server 14 determines the network devices connected to the control device 13-1 (control device ID = S1) and the control device 13-2 (control device ID = S2) based on the statistical information. Calculate the total estimated load status.

  The management server 14 is connected to the network devices corresponding to the network device IDs = PS1, PS2, PS3,... For the control device 13-1 having the control device ID = S1, and the load status of these network devices. The total sum of prospects is 220 Mbps. Therefore, since the sum of expected load conditions is predicted to exceed the load distribution target threshold = 200 Mbps, the control device 13-1 is a target for distributing the load to other control devices, that is, the load distribution source control device. Is determined.

  On the other hand, the management server 14 is connected to the network device corresponding to the connected network device ID = PS4, PS5, PS6,... When the sum of the estimated load status by the device is calculated, it is 31 Mbps. Accordingly, it is predicted that the sum of the expected load conditions is less than the stable operation threshold value = 50 Mbps, and therefore, it is determined that the load distribution destination control device is the target.

  Then, the management server 14 determines whether or not to perform load distribution and whether or not to become a control device candidate of the load distribution destination for all the control devices (in the case of FIG. 1, the control devices 13-1 and 13-2). To do.

Then, for example, load distribution control is performed in which management of a network device connected to the control device 13-1 that is determined to require load distribution is transferred to, for example, the control device 13-2 that is less than the stable operation threshold.
In this case, the priority of the control device with a lower expected load status is set higher, and the control device having the higher priority is selected as the load distribution destination control device.

  In the example of FIG. 7, it is possible to calculate whether the load on each network device is increasing or decreasing based on past statistical information and include it in the determination basis.

Specifically, in the example of FIG. 7, the increase tendency ratio = 60% in the control device 13-1 and the increase tendency ratio = 20% in the control device 13-2.
Therefore, if the control device 13-1 is also a target candidate for the load balancing destination control device, the control device 13-2 has a higher priority than the control device 13-1, so the control device 13-1 The device 13-2 is a load balancing destination control device.

  In this way, for a control device with a large ratio of network devices in which the load tends to increase, the priority as a load balancing destination control device is set low, and for a control device with a small increase tendency ratio By preferentially setting the load distribution destination, the load can be distributed more uniformly.

FIG. 8 is an explanatory diagram of an example of a connection destination changing operation in load distribution.
When the management server 14 determines that the load balancing is necessary for the control device 13-1, and determines that the load balancing destination control device is the control device 13-2, the management server 14 sends the load balancing destination to the control device 13-1. A load distribution instruction C1 including device information is transmitted.

  Based on the information on the load distribution destination device included in the received load distribution instruction C1, the control device 13-1 instructs the network device 11-1 connected to itself to change the connection destination. Thereby, the network device 11-1 instructed to change the connection destination changes the next call connection destination to the control device 13-2 and performs call connection processing.

FIG. 9 is an explanatory diagram illustrating an example of a case where the control device that has received the load distribution instruction selects a network device that changes the connection destination.
FIG. 9 shows a case where the control device 13-1 corresponding to the control device ID = S1 changes the connection destination.

  When the control apparatus 13-1 selects a connection destination change target apparatus based on the call information CI held by itself, the control apparatus 13-1 preferentially selects a network device having a large load status or a network device that tends to increase the load. . When the management server calculates and includes the number of network devices to be changed in the connection destination in the load distribution instruction, the target device is selected until the number reaches the specified number in order of priority.

In the case of the example of FIG. 9, the number of network devices to be connected is changed = 2. In this case, the connection destinations of the two network devices having the network device ID = PS1 and PS2 are changed. The Rukoto.
If the number of network devices is not included in the load distribution instruction, it is determined autonomously according to the load status of the control device itself.

FIG. 10 is an explanatory diagram of an example of a communication sequence related to load distribution.
In FIG. 10, the network device 11-1 with the network device ID = PS1 and the network device 11-2 with the network device ID = PS2 are communicating with the control device 13-1 with the control device ID = S1 (steps ST11 and ST12). ).

Thereby, the control apparatus 13-1 accumulate | stores the call information CI for every connected network device 11-1, 11-2 (step ST13).
The network device 11-3 with the network device ID = PS3 is communicating with the control device 13-2 with the control device ID = S2 (step ST14).

  Thereby, the control apparatus 13-2 accumulate | stores the call information CI of the connected network device 11-3 (step ST15).

On the other hand, when the management server 14 arrives at a call information CI collection opportunity at regular intervals (step ST16), the management server 14 sends a call information CI acquisition request to the control devices 13-1 and 13-2 (step ST17).
Thereby, the control device 13-1 and the control device 13-2 that have received the call information CI acquisition request from the management server 14 notify the management server 14 of the call information CI accumulated for a certain period of time as the call information CI notification. (Step ST18).

The management server 14 performs load distribution determination based on the collected call information CI (step ST19).
The management server 14 identifies the control device of the load distribution destination device and sends a load distribution instruction to the control device that is determined to require load distribution as a result of the load distribution determination in step ST19 (step ST20). . Specifically, in the case of the example in FIG. 10, the management server 14 distributes the control device 13-2 to the control device 13-1 (load distribution source control device) determined to require load distribution. A load distribution instruction for the destination device is transmitted.

  Receiving the load distribution instruction from the management server 14, the control device 13-1 selects a network device for switching the connection destination (step ST21), and sends a connection destination switching instruction to the selected network device (step ST22). Specifically, in the example of FIG. 10, the control device 13-1 selects the network device 11-1 as a network device for switching the connection destination, and controls the connection switching destination for the selected network device 11-1. A connection destination switching instruction for the device 13-2 is transmitted.

Receiving the connection destination change instruction from the control device 13-1, the network device 11-1 updates the control device information to which it is connected from the control device 13-1 to the control device 13-2 (step ST23).
Then, at the next call connection, the network device 11-1 transmits a connection request to the control device 13-2 to perform call connection processing (step ST24).

As a result, as a result, the control device 13-2 makes a connection response in response to the connection request from the network device 11-1 (step ST25).
Therefore, after that, the network device 11-1 is under the control of the control device 13-2 and performs communication (step ST26).

As described above, according to the first embodiment, the management server periodically collects session information of the control device for the purpose of appropriately distributing the access load from the network device to the control device, Statistical trend is grasped based on session information of each control device.
Therefore, by more effectively determining the load balancing source control device, load balancing timing, and load balancing destination control device, it is possible to perform dynamic load balancing according to the geographical and temporal environment of the control device. Become.

[2] Second Embodiment The second embodiment is an embodiment in the case where load distribution is performed based on a geographical environment of whether a network device is used in a business district or a residential district. .
In the second embodiment, the system configuration, the device configuration, the call information CI collection method, and the like are the same as those in the first embodiment, and a detailed description thereof will be omitted.

FIG. 11 is an explanatory diagram of an example in which the geographical environment of the network device is estimated based on the call information.
The management server 14 estimates the environment in which the network device is placed on the basis of the past temporal statistical information regarding the throughput for each network device.

  For example, as shown in FIG. 11A, the network device 11-1 has a low load in the morning and night and a large load in the daytime as a time trend of the throughput. Such a network device 11-1 is assumed to be installed in the “business district”, and the estimated geographical environment information is added to the management information of the network device.

  As shown in FIG. 11B, the network device 11-2 has a small daytime load and a high morning and night load as a time trend of the throughput. It is assumed that such a network device is installed in a residential area, and the estimated geographical environment information is added to the management information of the network device.

In the example of FIG. 11, the geographical environment is estimated using the throughput as an index. However, the geographical environment can be estimated based on the access frequency.
In addition, the geographical environment to be estimated is not limited to business districts / residential districts, but the location where the network device is installed is in the city center or suburbs, or is not strictly location information, but is frequently used for voice communication or data communication It is also possible to add information such as whether the area is frequent or not as geographical environment information.

FIG. 12 is an explanatory diagram of an example in the case of determining load distribution based on the load status of the control device using the throughput and the geographical environment as indices.
The management server 14 calculates the expected load status of each of the network devices 11-1 and 11-2 in the load distribution processing operation. Then, based on the calculation result of the expected load status, the control device (13-1, 13-2) in which the sum of the expected load status of the connected network devices exceeds the load distribution target threshold is distributed to the other control devices. It is targeted for.

In addition, control devices (13-1, 13-2) whose estimated total load status of the network devices to be connected is less than the stable operation threshold are set as candidates for load balancing destination control devices.
In the case of the example in FIG. 12, the load distribution target threshold is 200 Mbps, and the stable operation threshold is 50 Mbps.
More specifically, the management server 14 determines the network devices connected to the control device 13-1 (control device ID = S1) and the control device 13-2 (control device ID = S2) based on the statistical information. Calculate the total estimated load status.

  The management server 14 is connected to the network devices corresponding to the network device IDs = PS1, PS2, PS3,... For the control device 13-1 having the control device ID = S1, and the load status of these network devices. The total sum of prospects is 220 Mbps. Therefore, since it is predicted that the sum of expected load conditions exceeds the load distribution target threshold = 200 Mbps, it is determined that the control device 13-1 is a load distribution source control device.

On the other hand, the management server 14 is connected to the network device corresponding to the connected network device ID = PS4, PS5, PS6,... When the sum of the estimated load status by the device is calculated, it is 31 Mbps. Accordingly, it is predicted that the sum of the expected load conditions is less than the stable operation threshold value = 50 Mbps, and therefore, it is determined that the load distribution destination control device is the target.
Then, whether the management server 14 becomes a necessity candidate for load distribution and a candidate for a load distribution destination control device for all the control devices (in the case of FIG. 1, the control devices 13-1 and 13-2). Determine.

Then, for example, load distribution control is performed in which management of a network device connected to the control device 13-1 that is determined to require load distribution is transferred to, for example, the control device 13-2 that is less than the stable operation threshold.
In this case, the priority of the control device with a lower expected load status is set higher, and the control device having the higher priority is selected as the load distribution destination control device.

  In the example of FIG. 12, it is possible to calculate the geographical environment information of the network device and include it in the determination basis. For the control device 13-1 having a high ratio of network devices installed in the business district, the priority as the control device of the load distribution destination in the day load distribution is set low, and the priority in the night load distribution is set. Set high.

  On the other hand, for the control device 13-2 having a high ratio of network devices installed in a residential area, that is, for the control apparatus 13-2 having a low ratio of network devices installed in a business district, the daytime load distribution is performed. The priority as a load balancing destination control device is set high, and the priority for night load balancing is set low.

  Specifically, in the case of the example of FIG. 12, in the control device 13-1, the ratio of network devices installed in the business district is 60%, and in the control device 13-2, the network device is installed in the business district. Network device ratio = 20%.

Therefore, if the control device 13-1 is also a candidate for a load balancing destination control device, the control device 13-2 is compared with the control device 13-1 during the daytime period. Since the priority is higher, the control device 13-2 becomes the load balancing destination control device.
On the other hand, in the night time zone, the control device 13-2 has a lower priority than the control device 13-1, so that the control device 13-1 is a load balancing destination control device.

  As described above, according to the second embodiment, for a control device having a high ratio of network devices installed in a business district, the priority as a control device as a load distribution destination in the daytime time zone. The control device to which the load is distributed preferentially for a control device having a low ratio and a high ratio of network devices installed in a residential area, that is, a low ratio of network devices installed in a business district By doing so, it becomes possible to distribute the load more evenly according to the time zone.

[3] Third Embodiment Each of the above embodiments did not consider the call type. However, in the third embodiment, the load distribution process is performed based on the call type. It is an embodiment.
That is, a network device having a high frequency of voice communication tends to have a longer communication time for one session than a network device having a high frequency of data communication, and is not suitable for switching connection destinations.

Therefore, the priority as the target candidate of the control device of the load distribution destination is low for the network device having a high frequency of voice communication, and the priority as the target candidate of the control device of the load distribution destination is set for the network device having a high frequency of data communication. Configure to set higher.
As a result, more effective load distribution can be achieved. In this case, load distribution can be realized by including network device information for changing the connection destination in the load distribution instruction from the management server 14.

[4] Fourth Embodiment In the first embodiment, the connection destination of one network device is switched by one connection destination switching instruction. However, in the fourth embodiment, one connection destination switching instruction is performed. This is an embodiment when switching the connection destinations of a plurality of network devices.
FIG. 13 is an explanatory diagram of an example of a connection destination changing operation in load distribution according to the fourth embodiment.
When the management server 14 determines that load distribution is necessary for the control device 13-1, the load distribution destination control device is calculated as the control device 13-2, and the connection destination is changed based on the expected load status. A plurality of network devices 11-1 and 11-2 to be performed are selected, and information specifying these network devices is included in the load distribution instruction and transmitted.

  Accordingly, the control device 13-1 instructs the corresponding network device to change the connection destination based on the received information on the load distribution destination device and the network device information for changing the connection destination. The network devices 11-1 and 11-2 instructed to change the connection destination change the next call connection destination to the control device 13-2 and perform call connection processing.

A specific communication procedure will be described below.
FIG. 14 is an explanatory diagram of an example of a communication sequence related to load distribution in the fourth embodiment.
In FIG. 14, the network device 11-1 with the network device ID = PS1 and the network device 11-2 with the network device ID = PS2 are communicating with the control device 13-1 with the control device ID = S1 (steps ST31 and ST32). ).

Thereby, the control apparatus 13-1 accumulate | stores the call information CI for every connected network device 11-1 and 11-2 (step ST33).
Further, the network device 11-3 with the network device ID = PS3 is communicating with the control device 13-2 with the control device ID = S2 (step ST34).

  Thereby, the control apparatus 13-2 accumulate | stores the call information CI of the connected network device 11-3 (step ST35).

On the other hand, when the management server 14 arrives at a call information CI collection opportunity at regular intervals (step ST36), the management server 14 sends a call information CI acquisition request to the control devices 13-1 and 13-2 (step ST37).
Thereby, the control device 13-1 and the control device 13-2 that have received the call information CI acquisition request from the management server 14 notify the management server 14 of the call information CI accumulated for a certain period of time as the call information CI notification. (Step ST38).

The management server 14 performs load distribution determination based on the collected call information CI (step ST39).
The management server 14 identifies the control device of the load distribution destination device and a plurality of network devices that are the targets of load distribution for the control device determined to be load distribution as a result of the load distribution determination in step ST39. The load distribution instruction is transmitted (step ST40). Specifically, in the example of FIG. 14, the management server 14 sets the control device 13-2 as the load distribution destination device for the control device 13-1 that is determined to require load distribution, and performs load distribution. A load distribution instruction specifying a plurality of target network devices 11-1 and 11-2 is transmitted.

  The control device 13-1 that has received the load distribution instruction from the management server 14 transmits a connection destination switching instruction to the network devices 11-1 and 11-2 instructed to switch the connection destination (step ST41). Specifically, in the case of the example in FIG. 14, the control device 13-1 sets the connection switching destination for the network devices 11-1 and 11-2 designated as the network device for switching the connection destination. The connection destination switching instruction is sent out.

The network device 11-1 that has received the connection destination change instruction from the control device 13-1 updates the control device information to which it is connected from the control device 13-1 to the control device 13-2 (step ST42).
Similarly, the network device 11-2 that has received the connection destination change instruction from the control device 13-1 updates the control device information to which it is connected from the control device 13-1 to the control device 13-2 (step ST43).

Then, the network device 11-1 transmits a connection request to the control device 13-2 at the next call connection to perform a call connection process (step ST44).
As a result, the control device 13-2 makes a connection response in response to the connection request from the network device 11-1 (step ST45), and thereafter, the network device 11-1 manages the control device 13-2. Communication is carried out (step ST46).

Similarly, at the next call connection, the network device 11-2 transmits a connection request to the control device 13-2 to perform call connection processing (step ST47).
As a result, the control device 13-2 makes a connection response in response to the connection request from the network device 11-2 (step ST48), and thereafter, the network device 11-2 manages the control device 13-2. Communication is carried out (step ST49).

  As described above, according to the fourth embodiment, the control devices connected to the plurality of network devices 11-1 and 11-2 are reliably switched by a single load distribution instruction from the management server 14. And load distribution can be performed more easily.

[5] Fifth Embodiment The fifth embodiment is an embodiment in the case where load distribution is determined based on the load status of the control apparatus using the throughput, the temporal trend of the throughput, and the geographical environment as an index. .
FIG. 15 is an explanatory diagram of an example in the case of determining load distribution based on the load status of the control device using the throughput, the temporal trend of the throughput, and the geographical environment as an index.

  The management server 14 calculates the estimated load status of each network device in the load distribution processing operation, and sets the control device in which the sum of the estimated load status of the connected network devices exceeds the load distribution target threshold as the load distribution target.

In addition, a control device having a total load status estimate of a network device to be connected is less than the stable operation threshold is set as a candidate for a load balancing destination control device.
FIG. 15 shows an example in which the load distribution target threshold is 200 Mbps and the stable operation threshold is 50 Mbps.

  Based on the statistical information, the sum of the expected load statuses of the network devices corresponding to the network devices ID = PS1, PS2, PS3,... Connected to the control device 13-1 corresponding to the control device ID = S1 is calculated, and the load distribution is performed. Since 220 Mbps exceeding the target threshold is predicted, it is determined that load balancing needs to be performed for the control device 13-1.

  Similarly, the sum of the expected load conditions of the network device ID = PS4, PS5, PS6,... Connected to the control device 13-2 corresponding to the control device ID = S2 is calculated, and the control device corresponding to the control device ID = S3. The sum of the expected load status of the network device IDs connected to 13-3 = PS7, PS8, PS9,... Is calculated, and the network device IDs connected to the control device 13-4 corresponding to the control device ID = S4 = PS10, PS11. , PS 12,...

In the case of the example in FIG. 15, the expected load status of the control devices 13-2, 13-3, and 13-4 is predicted to be below the stable operation threshold of 50 Mbps. 13-4 are all candidates for the load balancing destination control device.
The management server 14 calculates the necessity of load distribution and the load distribution destination control device for all the control devices 13-1 to 13-4, and determines that the load distribution needs to be performed. For example, load distribution control is performed in which a network device connected to 1 (control device of load distribution source) is transferred to, for example, a control device 13-2 that is less than the stable operation threshold.

  In this case, the management server 14 increases the priority of the control device having a low load status expectation, and selects the control device having a high priority as the load distribution destination control device. Further, in the example of FIG. 15, the network device increase tendency ratio and the geographical environment information are combined and included in the determination basis of the load balancing destination control device. Here, in the increasing tendency ratio and the geographical environment information, since the increasing tendency ratio is considered to be more reliable with respect to the temporal tendency, the determination of the increasing tendency ratio is given priority.

In addition, when all of the control devices 13-2, 13-3, and 13-4 indicate the stable operation threshold value or less and the difference in the expected load situation is small, the increase tendency ratio of the network devices to be connected is determined.
In the example of FIG. 15, the control device 13-3 having an increase tendency ratio threshold value of 50% and higher than the increase tendency threshold value is excluded from the target candidates of the load distribution destination control device.
On the other hand, for the control devices 13-2 and 13-4 indicating the increase tendency threshold value or less, the business district threshold value that is the geographical environment information is subsequently determined.

  In the example of FIG. 15, when the business district threshold is 50% and the load distribution execution timing is daytime, the control device 13-4 in which the business district ratio of the connected network devices is higher than the business district threshold is the load distribution destination. Excluded from target control device candidates.

In this way, the management server 14 determines the control device 13-2 that has received the control device 13-2 as the transfer destination control device 13-2, and performs load distribution.
Further, in the fifth embodiment, when the target candidates of the load balancing destination control device are evaluated to be the same on any determination basis, the control device with the smallest load prediction probability is selected.

  As described above, according to the fifth embodiment, as in the first embodiment, the statistical trend is grasped based on the session information of each control device, and the load balancing source control device, load By determining the distribution timing and the control device of the load distribution destination, it becomes possible to more effectively perform dynamic load distribution according to the geographical and temporal environment of the control device.

[3] Modification of Embodiment The control devices 13-1 to 13-4 and the management server 14 according to this embodiment include a control device (control unit) such as an MPU and a storage device (memory unit) such as a ROM and a RAM. , An external storage device such as an HDD and a CD drive device, a display device such as a display device, and an input device such as a keyboard and a mouse, and has a hardware configuration using a normal computer.

  The control program executed by the control devices 13-1 to 13-4 and the management server 14 according to the present embodiment is a file in an installable format or an executable format, and is a CD-ROM, flexible disk (FD), CD-R. And recorded on a computer-readable recording medium such as a DVD (Digital Versatile Disk).

Further, the control program executed by the control devices 13-1 to 13-4 and the management server 14 of the present embodiment is provided by being stored on a computer connected to a network such as the Internet and downloaded via the network. You may comprise as follows. Further, the control program executed by the centralized control apparatus of the present embodiment may be configured to be provided or distributed via a network such as the Internet.
In addition, the control programs of the control devices 13-1 to 13-4 and the management server 14 according to the present embodiment may be provided by being incorporated in advance in a ROM or the like.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 10 Communication system 11 Network device 12 Communication network 13 Control apparatus 14 Management server 21 Call connection process part 31 Call control part 32 Call information management part 33 Connection destination control part 34 Call information storage table 34A Network device ID data 34B Call information ID data 34C Date data 34D Communication time data 34E Call type data 34F Throughput data 34G Priority data 41 Call information acquisition unit 42 Geographic environment estimation unit 43 Time statistics calculation unit 44 Load distribution instruction unit 45 Database CI Call information C1 Load distribution instruction ES Estimated data GI Geographic environment information

Claims (16)

Multiple network devices,
A plurality of control devices that perform communication control of the network devices assigned to each of them via a communication network;
A plurality of the control devices are managed via the communication network, and each network device is installed based on the current load status accompanying the communication control of the control device and the temporal transition of the load on each network device. estimating a geographical environment with, on the basis of the load conditions of the future due to the communication control of the control device which is predicted by the estimated geographical environment, perform reassignment of pairs pre SL network device to the control device A management server that issues load balancing instructions to
A communication system comprising: The management server causes the network devices to be reassigned based on a ratio of network devices that is estimated that the geographical environment is a business district.
The communication system according to claim 1. The management server is configured to control each of the plurality of control devices based on a ratio of network devices that are estimated to have the geographical environment as the business district among all the network devices currently assigned to the plurality of control devices. A load distribution source control device to which the network device assigned to the device is assigned to another control device, and a load distribution destination to which the network device currently assigned to another control device is assigned Discriminate between the control device and
The communication system according to claim 1 or 2. The management server determines whether the network device currently assigned to each control device is different from the network devices currently assigned to each of the control devices based on a ratio of network devices whose load is increasing. A load balancing source control device to be assigned to the control device and a load balancing destination control device to which the network device currently assigned to another control device is assigned;
The communication system according to claim 3 . The management server has a ratio of network devices that are estimated to have an increasing trend among all the network devices currently allocated to each of the control devices, exceeding a predetermined increasing threshold. The control device is determined as the load balancing source control device,
The communication system according to claim 4. The management server determines the control device whose load is equal to or less than a predetermined stable operation threshold as a control device candidate that can be the load distribution destination control device.
The communication system according to claim 4 or 5. The management server sets the control device having a lower load among the control devices that are candidates for the control device as the control device having a higher priority as the load distribution destination control device.
The communication system according to claim 6. The management server, the time course of the load in the same time zone of the day before, the temporal change of the load in the same time zone of the same day of the week, or, in the time course of the load in the same time period in the previous same month last year Based on at least one of them , predicting the time course of future loads,
The communication system according to any one of claims 1 to 7.   The communication system according to any one of claims 1 to 8, wherein the load is a throughput. The geographical environment is either the business district or the residential district.
The communication system according to any one of claims 1 to 9 . The management server of the plurality of control devices, if the proportion of the network device exceeds a predetermined business district ratio threshold the geographic environment is estimated to be the business district, the controller Is determined as the load balancing source control device,
The communication system according to claim 10. The management server of the plurality of control devices, the ratio of network devices the geographical environment is estimated to be the business district wherein the controller is lower than a predetermined business district ratio threshold, the load It is determined as a control device candidate that is a candidate for the control device of the distribution destination.
The communication system according to claim 10 . The management server of the controller that is the control device candidates, a network device that the geographical environment to the number of all the network devices under communication control of each control device is estimated to be the business district The control device having a lower ratio is the control device candidate having a higher priority as the load balancing destination control device.
The communication system according to claim 12 . Estimating the geographical environment in which each of the network devices is located based on temporal changes on the same day of the load;
The communication system according to any one of claims 1 to 13 . A plurality of network devices, a plurality of control devices that perform communication control of the network devices assigned to each of them via a communication network, and a plurality of the control devices that are connected via the communication network A management server for managing
Estimating means for estimating a geographical environment in which each network device is installed based on a temporal transition of a load in each network device;
Prediction means for predicting a future load situation caused by the communication control of the control device by estimated the geographical environment,
Based on the current load of the communication control status and predicted status of the future load of the control device for each control device, load balancing instructions for causing reassignment of the network device Management means;
Management server with A plurality of network devices, a plurality of control devices that perform communication control of the network devices assigned to each of them via a communication network, and a plurality of the control devices that are connected via the communication network A control program for controlling a management server for managing a computer,
The computer,
Estimating means for estimating a geographical environment in which each network device is installed based on a temporal transition of a load in each network device;
Prediction means for predicting a future load situation caused by the communication control of the control device by estimated the geographical environment,
Based on the current load of the communication control status and predicted status of the future load of the control device for each control device, load balancing instructions for causing reassignment of the network device Management means;
Control program to function.

Images