JP7045129B2 - Session control server switching method, management server and program - Google Patents

Description

The present invention relates to a technique for switching a session control server (SIP (Session Initiation Protocol) server) in a mobile communication network.

FIG. 1 is a system configuration diagram of a mobile communication network in the prior art.

The mobile communication network is built on a virtualization platform by EPC (Evolved Packet Core) and IMS (IP Multimedia Subsystem). The UE (User Equipment, mobile mobile terminal) communicates with a plurality of eNBs (evolved Node B, base station) via radio.

The EPC is composed of MME (Mobility Management Entity), SGW (Serving-Gateway), PGW (Proxy-Gateway) and PCRF (Policy and Charging Rules Function) to accommodate a plurality of eNBs (for example, Non-Patent Document 1). reference).
The MME is a core network control device, and executes UE movement management, UE authentication linked with HSS (Home Subscriber Server), and IP (Internet Protocol) transmission route setting control.
The SGW executes transmission control of IP packets based on the control from the MME.
The PGW serves as a connection point with a PDN (external network) such as the Internet or IMS, and receives all IP packets addressed to the UE from the PDN.
The PCRF determines the IP packet transmission policy such as QoS (Quality of Service) and billing method for performing transmission quality control in PGW and SGW.
The UE communicates with the IMS via eNB, SGW and PGW).

IMS is a service control method standardized by 3GPP (3rd Generation Partnership Project), and realizes a next-generation mobile phone network or NGN (Next Generation Network). As a result, the fixed network and the mobile network will be integrated, and all terminals will be provided with all IP. IMS provides UEs with common multimedia communication services (eg, IP phones, videophones, IM (Instant Message), etc.) with different IP-CANs (IP Connectivity Access Network) (eg, non-patent documents). 2).

IMS deploys a plurality of session control servers (SIP servers) to allocate necessary network resources and control gates. SIP is a call control protocol in IP that negotiates IP addresses, port numbers, and encodings required for communication between terminals, and controls outgoing and incoming calls.

IMS mainly has the following three functions.
CSCF (Call Session Control Function): Session control function AS (Application Server): Application service control function HSS (Home Subscriber Server): User information storage function

As session control, CSCF executes setting / releasing of a communication session between UEs, selection of AS according to service conditions, and transfer of SIP signal and SDP (Session Description Protocol) signal.
CSCF is divided into the following three functional elements.
P-CSCF (Proxy-CSCF)
S-CSCF (Serving-CSCF)
I-CSCF (Interrogating-CSCF)
The P-CSCF is a server that sends and receives packets to and from the UE via an access network.
S-CSCF is a server that executes user registration processing in cooperation with HSS.
The I-CSCF is a server that transfers SIP signals to the appropriate S-CSCF.
According to FIG. 1, S-CSCF and I-CSCF are collectively represented as I / S-CSCF.

The P-CSCF is accessed from the UE via the PGW. The P-CSCF extracts media information for each session and instructs the gate control and resource control for the PGW when the session is established. The P-CSCF is determined at the time of location registration from the UE and establishes a secure IPsec tunnel with the PGW (or SIP-enabled UE).
Therefore, in IMS, the P-CSCF is redundantly configured, and when a failure occurs in the P-CSCF, it is immediately switched to another P-CSCF to improve the fault tolerance of the entire system.
Upon switching the P-CSCF, the UE needs to release the session for the failed P-CSCF, execute the registration sequence again, and reestablish the session for the other P-CSCF. Various user communication recovery techniques have been proposed for P-CSCF failures (see, for example, Non-Patent Document 3).

Further, according to FIG. 1, a DNS (Domain Name System) server is arranged so that the PGW can immediately discover the P-CSCF.
The UE sends a DHCPv6 Information-Request message to the DSN server requesting the SIP server IPv6 address list option. In response, the DNS server returns a DHCP Reply message containing a list of IPv6 addresses of one or more P-CSCFs that can be assigned to the UE.

Generally, when the I / S-CSCF receives a message from the calling P-CSCF and sends a message to the called P-CSCF, a failure occurs at the timing of an error or no response from the called P-CSCF. Is detected. The I / S-CSCF notifies the HSS of the address of the P-CSCF where the failure occurred. The HSS notifies the MME to switch the called party P-CSCF to another P-CSCF. The MME sends a session release notification to the UE, and the UE executes the registration sequence again.

However, at the timing of this failure detection, since the transmission of the message at the time of incoming call fails, it has a great influence on the user communication. Therefore, it is necessary to periodically execute the life-and-death monitoring sequence for a plurality of redundantly arranged P-CSCFs.

FIG. 2 is a sequence for switching P-CSCF in the prior art.

<0th step: S0>
First, the UE sends a "connection request" to the EPC. The connection request is transferred to the MME via the SGW via the eNB. In response, the MME sends an "approval request" to the HSS via the PGW. The HSS executes the authentication process of the UE, and after registering as a subscriber, returns an "approval response" to the PGW.
On the other hand, PGW sends a "P-CSCF address list request" to the DSN server. This is a SIP server IPv6 address list option using a DHCPv6 Information-Request message and is based on the FQDN (Fully Qualified Domain Name).
On the other hand, the DSN server responds to the PGW with the P-CSCF address list. According to FIG. 2, the P-CSCF address list includes P-CSCF # 1 addresses and P-CSCF # 2 addresses.
The PGW returns to the UE a "connection permission" indicating permission to establish a bearer.
Then, the UE transmits an "approval response" to the PGW.

<First step: S1>
The UE then sends a "Register" to the PGW. The Register message contains a user ID called IMPI (IMS Private User Identity). IMPI is a unique identifier within IMS and is expressed in the same format as SIP URI.
On the other hand, PGW transfers the "registration request" to, for example, P-CSCF # 1. P-CSCF # 1 is based on any address included in the P-CSCF address list.
P-CSCF # 1 transfers the registration request received from PGW to I / S-CSCF as it is.
The I / S-CSCF sends a registration message including a user ID and a P-CSCF # 1 address to the HSS and requests an authentication process. In response, the HSS responds to the I / S-CSCF of the successful authentication. Then, the I / S-CSCF responds to P-CSCF # 1 with 200 OK as a success response.
P-CSCF # 1 responds to PGW with the received 200 OK.
The PGW responds to the UE with the received 200 OK.

<Second step: S2>
Next, the PGW periodically sends a "life-and-death monitoring message" to one or more P-CSCFs included in the P-CSCF address list received in S0. PGW itself monitors the occurrence of P-CSCF failures. The PGW that has transmitted the alive monitoring message detects that a failure has occurred in the P-CSCF if no error response is received or received from the P-CSCF.

Similarly, the DSN server 7 periodically sends a "life and death monitoring message" to one or more P-CSCFs. The DNS server also monitors the occurrence of P-CSCF failures. The DSN server that sent the alive monitoring message detects that a failure has occurred in the P-CSCF if no error response is received or received from the P-CSCF. At this time, the DSN server transmits a failure detection message including the address of the P-CSCF that has detected that the failure has occurred to the PGW.

Thereby, the PGW identifies the failed P-CSCF.

<Third step: S3>
Next, the PGW sends a "session release notification" to the UE.
This causes the UE to release a session with, for example, P-CSCF # 1 in which a failure has occurred. Specifically, the PGW may notify the UE of the address of the normal P-CSCF.

Then, the UE executes the same registration sequence as S1 described above, and sends a "Register" to the PGW in order to reestablish the session.
The PGW sends a registration request to another normal P-CSCF # 2 in order to identify the failure-occurring P-CSCF # 1.

As mentioned above, P-CSCF in IMS needs to switch from the failed P-CSCF to the alternative P-CSCF in order to restore the user communication with the UE in order to establish the session as the security of IPsec. There is. At this time, all UEs connected to the P-CSCF where the failure has occurred need to re-execute the registration sequence with the alternative P-CSCF.

3GPP TS 23.401, “General Packet Radio Service (GPRS) enhancements for Evolved Universal Terrestrial Radio Access Network (E-UTRAN) access”. 3GPP TS 23.228, “IP Multimedia Subsystem (IMS); Stage 2”. 3GPP, TS 23.380, “IMS Restoration Procedures”.

According to S2 in FIG. 2 described above, PGW regularly monitors a plurality of P-CSCFs for life and death. Then, when the PGW detects a failure of the P-CSCF, it is necessary to send a session release notification to the UE in order to execute the accommodation transfer to another P-CSCF. At this time, it is necessary to separately construct a DSN server in order to distribute the address list of the normal P-CSCF to be the accommodation replacement destination to the PGW.

Further, according to FIG. 2 described above, it is necessary to execute life-and-death monitoring from a large number of PGWs arranged in the EPC to a plurality of P-CSCFs for which a session has been established via the PGW. On the other hand, it is necessary to execute the life-and-death monitoring of P-CSCF from the DNS server located in IMS.
As a result, alive monitoring is executed for a plurality of P-CSCFs from a large number of PGWs and DNS servers, which is redundant and also causes an increase in the processing load of the entire EPC and IMS systems. ..

Therefore, the present invention provides a method, a management server, and a program that can smoothly switch the session control server when a failure occurs without making the life-and-death monitoring sequence of the session control server (P-CSCF) redundant. The purpose is to do.

According to the present invention
A gateway that is located in the EPC (Evolved Packet Core) and communicates with the user terminal, and a plurality of first session control servers that are located in the IMS (IP Multimedia Subsystem) and communicate with the gateway, and communicate with the first session control server. The second session control server and
It is a method of switching the session control server in the system having
The system further deploys a management server for each gateway that manages one or more connectable first session control server addresses.
The gateway sends a first session control server address list request to the management server, and the management server responds to the gateway with a list of addresses of the first session control server to which the gateway can connect.
The second session control server that received the registration request transmitted by the user terminal via the gateway and the first session control server transmits the registration request to the management server, and the management server is included in the registration request. The first step of associating and storing the user identifier and the first session control server address, and
A second step in which the management server detects a failure or a temporary failure of the first session control server and deletes the address of the first session control server that has detected the failure or the temporary failure from the list .
The management server sends a re-registration request including a user identifier associated with the first session control server in which a failure or a temporary failure is detected to the second session control server, and the second session control server determines. It is characterized by having a third step of executing a restore sequence in the first session control server.

According to another embodiment of the session control server switching method of the present invention.
The first session control server is a P-CSCF (Proxy --Call Session Control Function) server.
The second session control server is an S-CSCF (Serving --Call Session Control Function) server.
The gateway is PGW (Proxy -Gateway),
The management server is also preferably a P-CSCF management server.

According to another embodiment of the session control server switching method of the present invention.
About the first step
The user terminal sends a registration request to the gateway and
The gateway sends a registration request containing the user identifier and the first session control server address to the first session control server.
The first session control server transfers the registration request to the second session control server.
It is also preferable that the second session control server transfers the registration request to the management server.

According to another embodiment of the session control server switching method of the present invention.
About the second step
It is also preferable that the management server detects a failure by periodically transmitting a life-and-death monitoring message to each of the first session control servers.

According to another embodiment of the session control server switching method of the present invention.
About the second step
It is also preferable that the management server detects a temporary failure for any first session control server by using an operation command.

According to another embodiment of the session control server switching method of the present invention.
For the second step, the management server holds the address of the gateway that finally received the first session control server address list request when it detects a failure or provisional failure.
The management server receives the first session control server address list request from the gateway based on the retained address, responds to the list, and then determines that the update of the list of the first session control server for all gateways is completed. It is also preferable.

According to another embodiment of the session control server switching method of the present invention.
About the third step
The second session control server sends the re-registration request received from the management server to the first session control server in which the failure or temporary failure is detected, and the first session control server causes a failure or temporary failure. It is also preferable to confirm that it is done.

According to another embodiment of the session control server switching method of the present invention.
The system further has a subscriber server capable of communicating with a gateway and a second session control server.
Prior to the first step, the gateway sends an approval request to the subscriber server when it receives a connection request from the user terminal.
Regarding the third step, the second session control server sends a restore request of the first session control server to the subscriber server, and the subscriber server sends a session release notification to the user terminal via the gateway. It is also preferable to do so.

According to the present invention
A system having a gateway located in the EPC and communicating with the user terminal, a plurality of first session control servers arranged in the IMS and communicating with the gateway, and a second session control server communicating with the first session control server. A management server that manages one or more connectable first session control server addresses for each gateway.
The registration request transmitted by the user terminal is received from the second session control server via the gateway and the first session control server, and the user identifier included in the registration request and the first session control server address are associated with each other. The management table to remember and
An address list transmission means that receives a first session control server address list request from the gateway and responds to the gateway with a list of addresses of the first session control server to which the gateway can connect.
For the first session control server, a life-and-death monitoring means that detects a failure or a temporary failure and deletes the address of the first session control server that has detected the failure or the temporary failure from the list .
It has a re-registration request transmitting means for transmitting a re-registration request including a user identifier associated with the first session control server in which a failure or a temporary failure is detected to the second session control server.
It is characterized in that the restore sequence in the first session control server is executed from the second session control server.

According to the present invention
A system having a gateway located in the EPC and communicating with the user terminal, a plurality of first session control servers arranged in the IMS and communicating with the gateway, and a second session control server communicating with the first session control server. A program that operates a computer installed in a management server that manages one or more connectable first session control server addresses for each gateway.
The registration request transmitted by the user terminal is received from the second session control server via the gateway and the first session control server, and the user identifier included in the registration request and the first session control server address are associated with each other. The management table to remember and
An address list transmission means that receives a first session control server address list request from the gateway and responds to the gateway with a list of addresses of the first session control server to which the gateway can connect.
For the first session control server, a life-and-death monitoring means that detects a failure or a temporary failure and deletes the address of the first session control server that has detected the failure or the temporary failure from the list .
The computer functions as a re-registration request transmission means for transmitting a re-registration request including a user identifier associated with the first session control server in which a failure or a temporary failure is detected to the second session control server.
It is characterized in that the restore sequence in the first session control server is executed from the second session control server.

According to the method, management server and program of the present invention, the session control server can be smoothly switched when a failure occurs without making the life-and-death monitoring sequence of the session control server (P-CSCF) redundant.

It is a system block diagram of the mobile communication network in the prior art. This is a sequence for switching P-CSCF in the prior art. It is a system block diagram of the mobile communication network in this invention. It is a sequence which switches P-CSCF in this invention. It is a sequence diagram of S2 in this invention. It is a sequence diagram of S3 in this invention. It is a functional block diagram of the P-CSCF management server in this invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 3 is a system configuration diagram of the mobile communication network in the present invention.

According to the system of FIG. 3, as compared with FIG. 1, a P-CSCF management server is arranged instead of the DSN server. The P-CSCF management server executes alive monitoring for a plurality of P-CSCF # 1 and # 2 and communicates with the I / S-CSCF.
P-CSCF: First session control server I / S-CSCF: Second session control server PGW: Gateway intervening between the UE and P-CSCF P-CSCF management server: Manages the address of P-CSCF Management server

FIG. 4 is a sequence diagram in the present invention.

According to the present invention, the following three steps are performed.
<S0> The PGW acquires the addresses of one or more connectable P-CSCFs (first session control servers) from the P-CSCF management server.
<S1> The UE sends a "registration request" and executes a registration sequence with the IMS. Further, the P-CSCF management server stores the user ID and the P-CSCF address included in the registration request in association with each other.
<S2> The P-CSCF management server detects a failure or a temporary failure for a plurality of P-CSCFs.
<S3> The P-CSCF management server transmits a re-registration request including a user ID associated with the P-CSCF that has detected a failure or a temporary failure to the I / S-CSCF (second session control server). .. Then, the I / S-CSCF executes the restore sequence in the P-CSCF.
In the following, each step will be described in detail.

<0th step: S0>
According to S0 of FIG. 4, as compared with S0 of FIG. 2, the PGW communicates with the P-CSCF management server instead of the DSN server.
First, the UE sends a "connection request" to the EPC. The connection request is transferred to the MME via the SGW via the eNB. In response, the MME sends an "approval request" to the HSS via the PGW. The HSS executes the authentication process of the UE, and after registering as a subscriber, returns an "approval response" to the PGW.
PGW sends a "P-CSCF address list request" to the P-CSCF management server. On the other hand, the P-CSCF management server responds to the PGW with a list of P-CSCF addresses.
Then, the PGW returns to the UE a "connection permission" indicating permission to establish a bearer.
Then, the UE transmits an "approval response" to the PGW.

<First step: S1>
According to S1 of FIG. 4, the exact same sequence as S1 of FIG. 2 (UE <-> PGW <-> P-CSCF # 1 <-> I / S-CSCF) is executed, and the following sequence is further executed. do.

The I / S-CSCF transfers the registration request (Register) received from P-CSCF # 1 to the P-CSCF management server as it is. According to the registration request, the user ID is specified by the SIP-URI included in the To header, and the P-CSCF # 1 address is specified by the address included in the Via header.
As a result, the P-CSCF management server is linked as follows.
User ID <-> P-CSCF # 1 address Then, the P-CSCF management server responds 200OK (success response) to I / S-CSCF.

<Second step: S2>
FIG. 5 is a sequence diagram of S2 in the present invention.

According to FIG. 5, PGW # 1, # 2, and # 3 periodically send a "P-CSCF address list request" (for example, for the P-CSCF server name pcscf1) to the P-CSCF management server. .. This is a sequence similar to S0, where the PGW periodically updates the list of normal P-CSCF addresses.
According to FIG. 5, the P-CSCF management server responds to a P-CSCF address list containing both P-CSCF # 1 and # 2 addresses to PGW # 1, # 2, and # 3, respectively.

[Life and death monitoring process for P-CSCF]
The P-CSCF management server sends a "life-and-death monitoring message" to each of the P-CSCFs to monitor the occurrence of P-CSCF failures. SIP OPTIONS as a life-and-death monitoring message is sent periodically.
The P-CSCF returns a normal response to the alive monitoring message received from the P-CSCF management server at the normal time. On the other hand, in the event of a failure, an error response is returned or no response is returned. As a result, the P-CSCF management server detects the occurrence of a P-CSCF failure.

[Temporary failure detection by operation command for specific P-CSCF]
It is assumed that the P-CSCF management server detects the occurrence of a temporary failure for a specific P-CSCF by using an operation command. The operation command is used when it is desired to intentionally release the accommodating user of a specific P-CSCF at the time of maintenance, failure recovery by the operator, or the like.

It is assumed that the P-CSCF management server detects the occurrence of a failure in, for example, P-CSCF # 1. At this time, the address of the P-CSCF # 1 in which the failure occurred is deleted (excluded) from the P-CSCF address list.
In addition, the "address of PGW # 1" that last requested the P-CSF address list is also retained.

After that, the P-CSCF management server receives the P-CSCF address list request from each of the PGWs, and responds to the updated P-CSCF address list.
As a result, each PGW updates its own cache with the received P-CSCF address list. The updated P-CSCF address list does not include the address of the failed P-CSCF # 1.
Then, according to FIG. 5, when the P-CSCF address list request is received from PGW # 1 by the pre-held PGW # 1 address and the updated P-CSCF address list is responded to it, all of them are received. It is determined that the update of the P-CSCF address list for PGW has been completed.

[Regarding the retention of the address of the PGW that last requested the P-CSF address list]
Before executing the P-CSCF switching sequence, it is necessary to complete the update (cache) of the P-CSCF address list for all PGWs connected to the failed P-CSCF # 1. Prior to that, if the P-CSCF switching sequence is executed, one of the PGWs may connect to the failed P-CSCF.
As a countermeasure, it is assumed that each PGW sends a P-CSCF address list request periodically (at the same time interval), and the P-CSCF management server immediately detects the occurrence of a failure in the P-CSF address list. The last requested "address of PGW # 1" is retained.

According to FIG. 5, the last PGW that has transmitted the P-CSCF address list request immediately before the occurrence of the failure of P-CSCF # 1 is PGW # 1. That is, next, when PGW # 1 sends a P-CSCF address list request, P-CSCF address list request is received from all other PGWs, and P-CSCF # 1 address is excluded from P-. The CSCF address list can be considered as answered.

Further, if a call is made to the UE accommodated by the P-CSCF # 1 in which the failure has occurred while waiting for the P-CSCF address list request from PGW # 1, the user communication will fail. .. This waiting time can be shortened by shortening the TTL (Time To Live) of the cache. However, by shortening the TTL, the update cycle of the P-CSCF address list (cache) is shortened, and the processing load on the PGW is increased, which is a trade-off.

<Third step: S3>
FIG. 6 is a sequence diagram of S3 in the present invention.

Next, the P-CSCF management server sends a "re-registration request" to the I / S-CSCF. The re-registration request is a SIP Request to which the re-registration request flag is added, and includes a user ID associated with the address of P-CSCF # 1 where the failure has occurred.
On the other hand, the I / S-CSCF transmits the re-registration request to the P-CSCF # 1 associated with the user ID of the re-registration request.
P-CSCF # 1 will not return a response to the I / S-CSCF if a failure has occurred, and will return an error response if it is in a temporary failure state due to an operation command. Thereby, the I / S-CSCF confirms that the P-CSCF # 1 has a failure or a temporary failure.
Then, the I / S-CSCF transmits a P-CSCF restore request (HSS based P-CSCF restoration) to the HSS.
The HSS sends a P-CSCF restore request to the MME and PGW.
The MME sends a "session release notification" to the UE.
As a result, the UE releases the session with P-CSCF # 1.

According to FIG. 6, after S3, S1 is executed again.
The UE again sends a "Register" to the MME.
On the other hand, the PGW transfers the "registration request" to, for example, P-CSCF # 2 included in the P-CSCF address list.
P-CSCF # 2 transfers the registration request received from PGW to I / S-CSCF as it is.
The I / S-CSCF sends a registration message including a user ID and a P-CSCF # 2 address to the HSS and requests an authentication process. In response, the HSS responds to the I / S-CSCF of the successful authentication. Then, the I / S-CSCF responds to P-CSCF # 2 with 200 OK as a success response.
P-CSCF # 2 responds to PGW with the received 200 OK.
The PGW responds to the UE with the received 200 OK.

The I / S-CSCF transfers the registration request (Register) received from P-CSCF # 2 to the P-CSCF management server as it is.
As a result, the P-CSCF management server is linked as follows.
User ID <-> P-CSCF # 2 address Then, the P-CSCF management server responds 200OK (success response) to I / S-CSCF.

FIG. 7 is a functional configuration diagram of the P-CSCF management server in the present invention.

According to FIG. 7, the P-CSCF management server has a P-CSCF address list transmission unit 10, a management table 110, a registration request reception unit 111, a life-and-death monitoring unit 12, and a re-registration request transmission unit 13. .. These functional components are realized by executing a program for operating the computer mounted on the P-CSCF management server.

[Management table 110]
The management table 110 stores the user ID and the P-CSCF address in association with each other.
User ID <-> P-CSCF address

[P-CSCF address list transmitter 10]
The P-CSCF address list transmission unit 10 receives the "P-CSCF address list request" from the PGW and responds to the PGW with the "P-CSCF address list" (see S2 and FIG. 5 described above). The P-CSCF address list includes the addresses of P-CSCF to which the PGW can be connected.

[Registration request receiving unit 111]
The registration request receiving unit 111 receives a "registration request" from the I / S-CSCF (see S1 and FIG. 4 described above). This registration request is transferred from the I / S-CSCF to the P-CSCF management server when the registration sequence is executed between the UE, the PGW, the P-CSCF, and the I / S-CSCF. The registration request includes a user ID and a P-CSCF address, which are output to the management table 110.

[Life and Life Monitoring Department 12]
The life-and-death monitoring unit 12 detects a failure or a temporary failure for a plurality of P-CSCFs (see S2 and FIG. 5 described above). Specifically, the life-and-death monitoring unit 12 periodically sends a "life-and-life monitoring message" to each of the P-CSCFs, and detects the occurrence of a failure when an error response is received or there is no response. The life-and-death monitoring unit 12 outputs the P-CCSF address where the failure has been detected to the re-registration request transmission unit 13.

[Re-registration request transmission unit 13]
The re-registration request transmission unit 13 transmits a "re-registration request" including a user ID associated with the P-CSCF in which a failure or a temporary failure is detected to the I / S-CSCF.
As a result, the restore sequence in the P-CSCF in which the failure or the temporary failure is detected is executed from the I / S-CSCF. After the UE is released from the session by the restore sequence, the UE executes the registration sequence again, so that the switching of the accommodation of the UE from the failed P-CSCF to the normal P-CSCF is completed.

As described in detail above, according to the method, management server and program of the present invention, the life-and-death monitoring sequence of the session control server (P-CSCF) does not become redundant, and session control is performed when a failure occurs. You can switch servers smoothly.

Various modifications, modifications and omissions of the above-mentioned various embodiments of the present invention within the scope of the technical idea and viewpoint of the present invention can be easily carried out by those skilled in the art. The above explanation is just an example and does not attempt to limit anything. The present invention is limited only to the scope of claims and their equivalents.

1 P-CSCF management server 10 P-CSCF address list transmitter 110 Management table 111 Registration request receiver 12 Life and death monitoring unit 13 Re-registration request transmitter 2, 21, 22 P-CSCF
3 I / S-CSCF
4 PGW
5 HSS
6 MME
7 eNB
8 UE
9 DSN server

Claims (10)

A gateway that is located in the EPC (Evolved Packet Core) and communicates with the user terminal, and a plurality of first session control servers that are located in the IMS (IP Multimedia Subsystem) and communicate with the gateway, and communicate with the first session control server. Second session control server to
It is a method of switching the session control server in the system having
The system further deploys, for each gateway, a management server that manages one or more connectable first session control server addresses.
The gateway sends a first session control server address list request to the management server, and the management server responds to the gateway with a list of addresses of the first session control server to which the gateway can connect. It is a thing
The second session control server that received the registration request transmitted by the user terminal via the gateway and the first session control server transmits the registration request to the management server, and the management server responds to the registration request. The first step of associating and storing the included user identifier and the first session control server address, and
A second step in which the management server detects a failure or a temporary failure of the first session control server and deletes the address of the first session control server that has detected the failure or the temporary failure from the list .
The management server sends a re-registration request including a user identifier associated with the first session control server in which a failure or a temporary failure is detected to the second session control server, and the second session control server causes the second session control server. , A method of switching a session control server, which comprises a third step of executing a restore sequence in the first session control server. The first session control server is a P-CSCF (Proxy --Call Session Control Function) server.
The second session control server is an S-CSCF (Serving --Call Session Control Function) server.
The gateway is PGW (Proxy-Gateway).
The method for switching a session control server according to claim 1, wherein the management server is a P-CSCF management server. About the first step
The user terminal sends a registration request to the gateway and
The gateway sends a registration request containing the user identifier and the first session control server address to the first session control server.
The first session control server transfers the registration request to the second session control server.
The method for switching a session control server according to claim 1 or 2, wherein the second session control server transfers the registration request to the management server. About the second step
The one according to any one of claims 1 to 3, wherein the management server periodically sends a life-and-death monitoring message to each of the first session control servers to detect a failure. How to switch the session control server. About the second step
The method for switching a session control server according to any one of claims 1 to 3, wherein the management server detects a temporary failure for an arbitrary first session control server by an operation command. Regarding the second step, when the management server detects a failure or a temporary failure, the management server holds the address of the gateway that finally received the first session control server address list request.
The management server receives the first session control server address list request from the gateway based on the retained address, responds to the list, and then completes the update of the list of the first session control server for all gateways. The method for switching a session control server according to any one of claims 1 to 5, wherein the determination is made. About the third step
The second session control server sends the re-registration request received from the management server to the first session control server in which the failure or temporary failure is detected, and the first session control server causes the failure or temporary failure. The method for switching a session control server according to any one of claims 1 to 6 , wherein it is confirmed that the occurrence has occurred. The system further comprises a subscriber server capable of communicating with the gateway and a second session control server.
In the first stage before the first step, when the gateway receives the connection request from the user terminal, it sends an approval request to the subscriber server.
Regarding the third step, the second session control server sends a restore request of the first session control server to the subscriber server, and the subscriber server releases the session to the user terminal via the gateway. The method for switching a session control server according to any one of claims 1 to 7 , wherein a notification is transmitted. It has a gateway that is arranged in the EPC and communicates with the user terminal, a plurality of first session control servers that are arranged in the IMS and communicate with the gateway, and a second session control server that communicates with the first session control server. A management server that is located in the system and manages one or more connectable first session control server addresses for each gateway.
The registration request transmitted by the user terminal is received from the second session control server via the gateway and the first session control server, and the user identifier included in the registration request and the first session control server address are associated with each other. The management table to remember and
An address list transmitting means that receives a first session control server address list request from the gateway and responds to the gateway with a list of addresses of the first session control server to which the gateway can connect.
With respect to the first session control server, a life-and-death monitoring means that detects a failure or a temporary failure and deletes the address of the first session control server that has detected the failure or the temporary failure from the list .
It has a re-registration request transmitting means for transmitting a re-registration request including a user identifier associated with the first session control server in which a failure or a temporary failure is detected to the second session control server.
A management server characterized in that a restore sequence in the first session control server is executed from the second session control server. It has a gateway that is arranged in the EPC and communicates with the user terminal, a plurality of first session control servers that are arranged in the IMS and communicate with the gateway, and a second session control server that communicates with the first session control server. A program that operates a computer installed in a management server that is located in the system and manages one or more connectable first session control server addresses for each gateway.
The registration request transmitted by the user terminal is received from the second session control server via the gateway and the first session control server, and the user identifier included in the registration request and the first session control server address are associated with each other. The management table to remember and
An address list transmitting means that receives a first session control server address list request from the gateway and responds to the gateway with a list of addresses of the first session control server to which the gateway can connect.
With respect to the first session control server, a life-and-death monitoring means that detects a failure or a temporary failure and deletes the address of the first session control server that has detected the failure or the temporary failure from the list .
The computer functions as a re-registration request transmission means for transmitting a re-registration request including a user identifier associated with the first session control server in which a failure or a temporary failure is detected to the second session control server.
A program for a management server, characterized in that a restore sequence in the first session control server is executed from the second session control server.

Images