JP2011188031A - REDUNDANT VoIP SYSTEM - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a redundant VoIP system for smoothly switching a system in a short time in a redundant structure. <P>SOLUTION: When a call processing section itself is of an active system, the call processing section stops call control according to the generation of a system switch request, and transmits a system switch start notice to a network interface. When the call processing section itself is of a new active system corresponding to the system switch start the notice, the call processing section captures a duplicate of call state information from a call processing section which if the active system in the past, resumes call control on the basis of the captured call state information, and transmits a system switch completion notice to the network interface. The network interface temporarily accumulates call processing packets according to the system switch start notice to stop a relay to the call processing section, and stops the accumulation of the call processing packets according to the system switch completion notice to resume the relay to the call processing section. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a redundant VoIP (Voice over IP) system having a redundant configuration consisting of at least two systems of an active system and a standby system, and more particularly to a new standby system for a previous active system without interrupting call processing. The present invention relates to a redundant VoIP system that switches a system configuration from a conventional standby system to a new operating system.

  Generally, in a redundant VoIP system, it is important not to make the communication partner aware of the occurrence of such a system switching request when switching between the active system and the standby system. For example, in the case of telephone communication, if there is a call in progress during system switching, “call continuation” is required so that the call can be continued without interrupting call processing for the call.

  Patent Document 1 discloses a system switching system capable of preventing a control signal from being lost and performing system switching while ensuring synchronization between systems during system switching between an active system and a standby system. .

JP 2007-334528 A

  However, in the prior art as disclosed in Patent Document 1, among events generated in response to a control signal in call processing, when an event waiting for a response from the outside still exists at the time of system switching, The system switching is suppressed until a response is received, and the system switching is completed when there is no response waiting event. However, this method has a disadvantage that the switching operation is waited until the event processing in the response waiting state is completed, and as a result, the time required for system switching becomes long. As the system switching time is prolonged, it is necessary to increase the buffer capacity to be prepared for accumulating voice data that arrives even during system switching. Furthermore, when the system is switched, the call processing unit itself may newly generate a control signal and create a new event waiting for a response, and there may be a case where the system switching cannot be performed smoothly.

  An object of the present invention is to provide a redundant VoIP system that achieves smooth system switching in a short time in a redundant configuration.

  The redundant VoIP system according to the present invention includes at least two call processing units that perform call control while updating call state information according to a call processing packet, and the call processing between the call processing unit and the IP network. A redundant VoIP system including at least one network interface unit that relays a packet, wherein the call processing unit responds to generation of a system switching request when the call processing unit is an active call processing unit. Then, the call control is stopped and a system switching start means for transmitting a system switching start notification to the network interface unit and a call processor of a new operating system corresponding to the system switching start notification Taking a copy of the call state information from a call processing unit that was previously active, resuming the call control based on the call state information, and sending a system switch end notification to the network A system switching end means for transmitting to the interface unit, wherein the network interface unit temporarily accumulates the call processing packet and stops relaying to the call processing unit in response to the system switching start notification The storage of the call processing packet is stopped and the relay to the call processing unit is restarted in response to the system switching end notification.

  The redundant VoIP system according to the present invention provides a redundant VoIP system that achieves smooth system switching in a short time.

It is a block diagram which shows the VoIP system in the Example of this invention. It is a figure which shows the structural example of a call processing packet and a switch notification packet. It is a figure which shows the structural example of call state information. 3 is a flowchart showing a processing procedure in a system switching operation of the call processing unit shown in FIG. 1. It is a block diagram which shows the VoIP system in the 1st modification of this invention. It is a block diagram which shows the VoIP system in the 2nd modification of this invention.

  Embodiments of the present invention will be described in detail with reference to the accompanying drawings.

<Example>
FIG. 1 shows a VoIP system 50 in an embodiment of the present invention. The VoIP system 50 is connected to an IP network 90, and a CA (Call Agent) server 91 is connected to the tip of the IP network 90. The VoIP system 50 exchanges a call processing packet P1 with the CA server 91, thereby making a call, an incoming call, and a call start for a call operation of a telephone terminal (not shown) connected to the IP network 90. And call control including call stop and billing. The call processing packet P1 is a packet that is transmitted and received in order to perform call control according to a protocol such as SIP (Session Initiation Protocol).

  In FIG. 1, a VoIP unit that converts a voice signal and an IP voice packet, a TDM switch unit that switches connection with a subscriber line, and a TDM interface that transmits and receives voice signals to and from the subscriber line. The illustration of the part is omitted for ease of explanation. Further, the description of the transmission / reception operation of the voice packet generated by the call is omitted.

  The VoIP system 50 includes a network interface unit 10 and two call processing units made redundant by having the same configuration, that is, a call processing unit 30 and a call processing unit 40. Each of the call processing unit 30 and the call processing unit 40 has the same internal function, and each provides a redundant function. Normally, either one is operated as an active system and the other is operated as a standby system.

  The network interface unit 10 relays call processing packets between the IP network 90 and the call processing unit 30 or 40. The network interface unit 10 includes an input processing unit 11, an output processing unit 12, an accumulation processing unit 13, and a packet multiple distribution unit 14. The input processing unit 11 is connected to the IP network 90 and receives a call processing packet P1 and a voice packet (not shown) from the CA server 91 via the IP network 90. The output processing unit 12 is connected to the IP network 90 and transmits a call processing packet P1 and a voice packet (not shown) to the CA server 91. The accumulation processing unit 13 temporarily accumulates the call processing packet P1, and sequentially transmits the packets to the packet multiplex distribution unit 14 using a FIFO (First In First Out) method. The accumulation processing unit 13 temporarily stops relaying to the packet multiplex distribution unit 14 in response to the accumulation instruction P3 from the packet multiplex distribution unit 14, and in response to the cancellation of the accumulation instruction P3. And a function of restarting relaying. The packet multiplex distribution unit 14 multiplexes the call processing packet P1 destined for the CA server 91 and distributes the call processing packet P1 from the CA server 91 to send the call processing packet P1 to the call processing unit. 30 or the call processing unit 40. The distribution process determines which of the call processing unit 30 and the call processing unit 40 is the active call processing unit, and performs the call processing to the active call processing unit (for example, the call processing unit 30). Packet P1 is transmitted.

  The call processing unit 30 includes an event monitoring unit 31, an event issuing unit 32, a protocol processing unit 33, a call state management unit 34, a call state takeover unit 35, and a system switching processing unit 36. Similarly, the call processing unit 40 includes an event monitoring unit 41, an event issuing unit 42, a protocol processing unit 43, a call state management unit 44, a call state takeover unit 45, and a system switching processing unit 46. The

  FIG. 1 shows two call processing units 30 or 40 for realizing two-system redundancy, but there is no limitation on the present invention, and there are three redundant systems according to the present invention. The above call processing unit may be included. Since the call processing unit 30 and the call processing unit 40 have the same functional configuration, their internal configurations will be described below using only the call processing unit 30 as an example.

  The event monitoring unit 31 of the call processing unit 30 receives the call processing packet P1 from the packet multiplex distribution unit 14 of the network interface unit 10 and notifies the protocol processing unit 33 of the occurrence of the event. The event issuing unit 32 detects the occurrence of an event corresponding to the process in the protocol processing unit 33, and transmits a call processing packet corresponding to the content of the event to the packet multiplex distribution unit 14 of the network interface unit 10. . The event issuing unit 32 also generates a switching notification packet P2 in response to the system switching instruction from the system switching processing unit 36 and transmits it to the network interface unit 10.

  The protocol processing unit 33 performs call control processing based on a call processing protocol (for example, SIP) in response to the occurrence of an event from the event monitoring unit 31, and calls status such as outgoing call, incoming call, call start / end, etc. Issuing events on the basis of The call state management unit 34 holds the call state of each telephone subscriber as call state information. The call state information is generated or changed as the call control process of the protocol processing unit 33 progresses.

  The call state takeover unit 35 synchronizes with another system in response to a system switching instruction from the system switching processing unit 36, and when it is a new active system at the time of system switching, the call state of the other system Take a copy of the information to yourself. On the other hand, when the system was previously active at the time of system switching, a copy of the call status information of the local system is provided to the other system. For example, at the time of system switching in which the call processing unit 30 is the previous active system and the call processing unit 40 is the new active system, the call state takeover unit 35 creates a copy of the call state information held by the call state management unit 34. . On the other hand, the call state takeover unit 45 takes in the copy from the call state takeover unit 35 of the other system and holds it in the call state management unit 44 of the own system.

  The system switching processing unit 36 detects the generation of a system switching request and instructs the call state takeover unit 35, the protocol processing unit 33, and the event issuing unit 32 to perform system switching. The system switching processing unit 36 cooperates with the system switching processing unit 46 of the other system to synchronize the timing of system switching and instruct system switching. Generation of the system switching request includes, for example, an input operation from the outside, occurrence of a failure, and arrival of a predetermined time.

  FIG. 2A shows a call processing packet P1 and a switch notification packet P2, respectively. The call processing packet P1 is composed of processing identification for specifying processing contents such as outgoing call, incoming call, call start / end, and call control information corresponding to the processing. The switch notification packet P2 includes a system switch identification that specifies an active system and a standby system, and a start flag and an end flag. The system switching identification includes, for example, information for identifying that the call processing unit 30 is an active system and the call processing unit 40 is a standby system. The start flag has an ON or OFF state. When the start flag is ON, the start of system switching is instructed. The end flag has an ON or OFF state. When the end flag is ON, the end of system switching is instructed.

  FIG. 2B shows a configuration example of the call state information P5. The call state information P5 includes a plurality of call state information tables T1 to Tn (n is a positive number) each corresponding to one call. One call state information table T1 includes, for example, call identification, calling subscriber number, called subscriber number, call status, call start time, call end time, voice compression method, and billing information assigned to each call. Etc. The call status may include not only status information such as outgoing call, incoming call, and call start / end, but also more detailed event occurrence status related to call control. The configuration and content of the call state information P5 may have various forms depending on the call control protocol used in the VoIP system 50, but the call can be continued by referring to the call state information P5 even during system switching. Information must be included.

  FIG. 3 shows the system switching operation of the call processing unit 30 and the call processing unit 40 for realizing call continuation. In this figure, the system state before switching is the call processing unit 30 for the active system and the call processing unit 40 for the standby system. Therefore, before the system switching, the call processing unit 30, which is the active system, is processing several calls, and the call state information P5 as shown in FIG. Held in the part 34. The call processing unit 40 which is a standby system is on standby and does not hold a call being processed. In the following description of the operation procedure, the constituent elements of each unit shown in FIG.

  First, the call processing unit 30 starts system switching when a system switching request is generated (step S1). The call processing unit 30 transmits a switching notification packet P2 to the network interface unit 10. Specifically, the system switching processing unit 36 instructs the protocol processing unit 33 and the event issuing unit 32 to perform system switching. In response to this instruction, the event issuing unit 32 generates a switch notification packet P2 and transmits it to the packet multiplex distribution unit 14 of the network interface unit 10. At this time, in the switching notification packet P2, the start flag is set to ON to indicate “system switching start”.

  Next, in response to the switch notification packet P2, the network interface unit 10 starts accumulating the call processing packet P1 received from the CA server 91 and stops distributing the call processing packet P1 (step S2). Specifically, the packet multiplex distribution unit 14 sets a storage instruction P3 to the storage processing unit 13 in response to the switching notification packet P2 indicating “system switching start”. After confirming the setting of the accumulation instruction P3, the accumulation processing unit 13 accumulates the call processing packet P1 received from the CA server 91, thereby temporarily stopping the packet relay to the packet multiplex distribution unit 14. Accordingly, all the call processing packets P1 received from the CA server 91 thereafter are temporarily stored in the storage processing unit 13. The packet multiplex distribution unit 14 does not distribute the call processing packet P1, and as a result, transmission of the call processing packet P1 to the call processing unit 30 or the call processing unit 40 is stopped.

  On the other hand, after transmitting the switching notification packet P2, the call processing unit 30 stops the response process and stops the generation of a new event (step S3). That is, the protocol processing unit 33 stops the call control regardless of whether there is a response waiting state waiting for a new call processing packet due to an already generated event, and stops the operation of issuing a new event itself. .

  Next, the call processing unit 30 determines whether or not it is in the “no processing state” (step S4). There may be a slight probability that the call processing packet P1 transmitted from the network interface unit 10 immediately before switching is still being processed. In this case, it is determined that there is no unprocessed state. In this case, the response process is stopped again after the process, and the generation of a new event is stopped (step S3). On the other hand, since the network interface unit 10 stops distributing the call processing packet P1, the call processing unit 30 becomes “no processing state” in a very short time, and is determined to be “no processing state”.

  Next, in response to the determination of “no processing state”, the call processing unit 30 transfers a copy of the call state information P5 toward the call processing unit 40 (step S5). Specifically, the call state takeover unit 35 of the call processing unit 30 creates a copy of the call state information P5 held in the call state management unit 34, and transfers this to the call state management unit 44 of the call processing unit 40. Forward towards. The call state information P5 includes call state information tables T1 to Tn as shown in FIG. 2B, and includes call state information of all calls being processed by the call processing unit 30 which was the active system at the time of switching. .

  On the other hand, the call processing unit 40 holds a copy of the call state information P5 in response to the transfer of the call state information P5 (step S6). A copy of the transferred call state information P5 is held in the call state management unit 44 of the call processing unit 40. Next, the call processing unit 40 ends the system switching (step S7). At this time, the call processing unit 40 transmits a switching notification packet P2 to the network interface unit 10. Specifically, the system switching processing unit 46 of the call processing unit 40 instructs the event issuing unit 42 to end the system switching, and the event issuing unit 42 sends the switching notification packet P2 to the packet multiplexing distribution of the network interface unit 10. To the unit 14. In the switch notification packet P2, the end flag is set to ON to indicate “system switch end”.

  In response to the switching notification packet P2, the network interface unit 10 stops the accumulation of the call processing packet P1, and resumes the distribution of the call processing packet P1 (step S8). Specifically, the packet multiplex distribution unit 14 of the network interface unit 10 cancels the setting of the accumulation instruction P3 for the accumulation processing unit 13 in response to the reception of the switch notification packet P2 indicating “system switch end”. The accumulation processing unit 13 resumes packet transmission to the packet multiplex distribution unit 14 in response to the cancellation of the setting of the accumulation instruction P3. The packet multiplex distribution unit 14 resumes the distribution of the call processing packet P1 toward the event monitoring unit 41 of the call processing unit 40 which is a new operational system.

  Next, the call processing unit 40 restarts the response process and the occurrence of a new event (step S9). Specifically, the protocol processing unit 43 of the call processing unit 40 recognizes an event that has already occurred based on the call state information held in the call state management unit 44 and restarts the response process, Issue new events as they progress. As a result of the above system switching operation, the system status after switching is that the active system is the call processing unit 40 and the standby system is the call processing unit 30.

  In the above embodiment, the call processing packet P1 is temporarily stored in the network interface unit 10 in the system switching between the call processing units 30 and 40. As a result, a no-process state can be created in the call processing unit 30 that was the active system immediately before switching, and system switching can be performed in an extremely short time without waiting for a response. In addition, a copy of the call state information P5 is transferred from the call processing unit 30 that was the active system immediately before switching to the call processing unit 40 that is newly the active system. As a result, the call processing unit 40 that newly becomes the active system takes over the response processing based on the call state information P5, and can continue the call smoothly in a short time.

  In this embodiment, the 1-system network interface section 10 is used for the 2-system call processing sections 30 and 40. In this case, the IP network 90 and the CA server 91 need not be aware of the system switching between the call processing unit 30 and the call processing unit 40. Therefore, the CA server 91 only needs to recognize the MAC address of the network interface unit 10 that is the other party in a fixed manner, and the CA server 91 changes the other party's MAC address when the system is switched on the VoIP system 50 side. There is no need, and there is no risk of packet loss due to the MAC address change.

  In addition, one floating (virtual) IP address can be set in advance for the plurality of call processing units 30 and 40, and the packet multiplex distribution unit 14 always processes packets to the floating IP address. By transmitting P1, an active call processing unit (for example, call processing unit 40) operating among the call processing units associated with the floating IP address can receive the processing packet P1. . As a result, it is not necessary to check the operation system in the packet multiplex distribution unit 14, and the system switching time can be further shortened.

  Further, in the synchronization process between the state takeover unit 35 and the state takeover unit 45, it is also possible to transfer a copy of the call state information P5 between the both when the call state changes, not at the time of system switching. As a result, the call state information P5 always matches in both systems, and it is not necessary to duplicate and transfer the call state information P5 at the time of system switching.

<First Modification>
FIG. 4 shows a VoIP system in the first modification of the present invention. In the above-described embodiment, the switching notification packet P2 transmitted and received according to the IP protocol is used for the system switching notification between the call processing units 30 and 40 and the network interface unit 10. On the other hand, as shown in the figure, the switching state signal P4 is used for the system switching notification in the first modification example, instead of the switching notification packet P2. The switching status signal P4 is transmitted from the system switching notification unit 37 of the call processing unit 30 to the system switching receiving unit 15 of the network interface unit 10, for example. The system switching receiving unit 15 sets the accumulation instruction P3 for the accumulation processing unit 13 in response to the switching notification by the switching state signal P4.

<Second Modification>
FIG. 5 shows a VoIP system in the second modification of the present invention. The VoIP system in FIG. 1 is composed of only the 1-system network interface unit 10. On the other hand, in the second modification, the network interface unit 10 shown in FIG. 1 has a redundant configuration, and consists of two systems of the first network interface unit 10 and the second network interface unit 20. In this case, one of the network interface units 10 and 20 is set as the active system, and the switching notification packet from the call processing units 30 and 40 is directed to the active network interface unit (for example, the network interface unit 20). System switching between the call processing units 30 and 40 is executed.

10, 20 Network interface unit 11, 21 Input processing unit 12, 22 Output processing unit 13, 23 Accumulation processing unit 14, 24 Packet multiple allocation unit 15 System switching reception unit 30, 40 Call processing unit 31, 41 Event monitoring unit 32, 42 Event issuing unit 33, 43 Protocol processing unit 34, 44 Call state management unit 35, 45 Call state takeover unit 36, 46 System switching processing unit 50 VoIP system 90 IP network 91 CA server P1 call processing packet P2 switching notification packet P3 accumulation instruction P4 switching state signal P5 call state information

Claims (5)

It comprises at least two call processing units that perform call control while updating call state information according to call processing packets, and at least one system that relays the call processing packets between the call processing unit and the IP network. A redundant VoIP system including a network interface unit,
The call processing unit includes:
When itself is an active call processing unit, in response to the occurrence of a system switching request, the call control is stopped, and system switching start means for transmitting a system switching start notification to the network interface unit;
If it is a new active call processor corresponding to the system switching start notification, it takes a copy of the call state information from the call processor previously active, and based on the call state information System switching termination means for resuming the call control and transmitting a system switching termination notification to the network interface unit,
The network interface unit temporarily stores the call processing packet in response to the system switching start notification and stops relaying to the call processing unit, and the call processing packet in response to the system switching end notification And the relay to the call processing unit is restarted.   The system switching start means stops the call control regardless of whether there is a response waiting state waiting for a new call processing packet due to an event that has already occurred by the call control, and the system switching end means The redundant VoIP system according to claim 1, wherein the call control is resumed by recognizing the already generated event based on the state information.   2. The redundant VoIP system according to claim 1, wherein the occurrence of the system switching request is at least one of an input operation, occurrence of a failure, and arrival of a predetermined time.   The redundant VoIP system according to claim 1, wherein the network interface unit relays the call processing packet only to an active call processing unit among the call processing units. The network switching unit includes at least two systems, and the system switching start unit and the system switching end unit notify the system switching start notification and the system switching end notification to an operating system network interface unit among the network interface units. The redundant VoIP system according to claim 1, wherein:

Images