PCF remote disaster recovery method

Abstract

The invention relates to a PCF remote disaster recovery method. The method comprises the following steps: step 1, deploying a multi-region PCF environment; step 2, after the PCF1 and the PCF2 are normally powered on and run to enter a working state, the PCF1 and the PCF2 are actively registered to the NRF through the NnrfNFManagementNFRegister, a heartbeat message between the PCF1 and the PCF2 and the NRF is periodically sent through the NnrfNFManagementNFUpdate according to a configuration period, and the NRF detects and maintains the working state of the PCF by using the heartbeat message; 3, under the abnormal condition, when the CDF1 breaks down due to faults, the state of the CDF1 is automatically monitored in a redis sentry mode; 4, the PCF2 judges whether the CDF3 is available or not, if the CDF3 works normally, the user information is obtained from the CDF3 and stored to the local PCF, and if the CDF3 cannot work normally, the CDF4 is connected, and the user information is obtained from the CDF4. The method has the advantages that the PCF is deployed in a plurality of areas, even if a fault occurs in one area, the system can still run continuously, and high availability of the system is guaranteed; 2, the data can be backed up in real time, the risk of data loss is reduced, and the integrity and reliability of the data are ensured;

Description

Technical Field

[0001] The present invention belongs to the field of mobile communication technology, and in particular relates to a PCF remote disaster recovery method. Background Art

[0002] At present, with the rapid development of mobile communication technology, China's mobile phone users have already exceeded one billion. Faced with increasingly fierce market competition, mobile operators will inevitably continue to improve service quality for users and continuously improve user experience. With the acceleration of 5G network deployment, various security issues and hidden dangers faced by the network are becoming more and more important. While paying attention to the performance of network elements, how to improve the reliability and security of network elements and reduce or avoid the impact on other network elements is the key to building a reliable 5G core network. The PCF (Policy Control Function, hereinafter referred to as PCF) is a very important core network element in the core network, providing mobility, UE access selection and PDU session-related policies for the AMF network element (Access and Mobility Management Function, hereinafter referred to as AMF) and the SMF network element (Session Management Function, hereinafter referred to as SMF). Therefore, the reliability of PCF is very important.

[0003] As the name implies, off-site disaster recovery is to establish a backup PCF in a different region. If the PCF fails, it is necessary to wait for the user to re-initiate an injection to the PCF before the business can be restored. The real-time takeover of the business cannot be achieved, which has a great impact on user perception and may cause users to be unable to communicate for a period of time. PCF off-site disaster recovery is a practice that ensures high availability and disaster recovery capabilities in different geographical regions. It involves synchronizing data and services between multiple geographical locations (usually different data centers or cloud regions) to ensure that even if a location fails, applications and services can continue to run in other locations. To achieve off-site disaster recovery for PCF, it is necessary to meet the requirements of multi-region deployment, data consistency and synchronization. PCF provides the ability to deploy across multiple geographical regions and supports dynamic switching of applications between multiple regions. By deploying PCF instances in multiple regions, applications can run in different data centers. In order to achieve disaster recovery, data needs to be synchronized between different regions. High data availability needs to be achieved through storage synchronization. Through this multi-region deployment method, it can run across multiple geographical locations. Once a region fails, the system can quickly switch to other regions. At the same time, each region of multi-region deployment is isolated from each other to prevent local failures from affecting the entire system. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the prior art in this technical field and provide a PCF remote disaster recovery method.

[0005] The PCF remote disaster recovery method comprises the following steps:

[0006] Step 1: Multi-region PCF environment deployment;

[0007] Step 2: After PCF1 and PCF2 are powered on and enter the working state, they actively register with NRF through Nnrf_NFManagement_NFRegister, and send heartbeat messages to NRF regularly through Nnrf_NFManagement_NFUpdate according to the configuration cycle. NRF uses heartbeat messages to detect and maintain the working state of PCF.

[0008] Step 3: In abnormal situations, when CDF1 fails, use the Redis sentinel mode to automatically monitor the status of CDF1;

[0009] Step 4, PCF2 determines whether CDF3 is available. If CDF3 works normally, the user information is obtained from CDF3 and saved locally in PCF. If CDF3 does not work normally, the user information is connected to CDF4 to obtain the user information from CDF4.

[0010] Furthermore, in step 1, the multi-region PCF environment deployment specifically includes:

[0011] Step 1.1, deploy a data center (DC);

[0012] Step 1.2: Select appropriate geographical locations to deploy a multi-region PCF environment based on business needs, and ensure that there is a stable network connection between regions to support data synchronization and failover. Deploy PCF1 and PCF2 in the two selected regions respectively, ensure that the regional environments can communicate with each other, and configure the same network security policies.

[0013] Step 1.3, build CDF1 (Centralized Data Function, hereinafter referred to as CDF) and CDF2 environment, CDF1 and CDF2 network elements are mutually active and standby, CDF1 is responsible for storing user online information in PCF1 network element;

[0014] Step 1.4, build CDF3 and CDF4 environment, CDF3 and CDF4 network elements are mutually active and standby, CDF3 is responsible for storing user online information in PCF2 network element;

[0015] Step 1.5: All CDFs are linked to DC. When CDF is used as the primary, it synchronizes data to DC. When CDF is used as the backup, it needs to synchronize user information from DC. Bidirectional synchronization is used between CDF and DC.

[0016] Step 1.6, configure the communication addresses and port numbers of CDF1, CDF2, CDF3, and CDF4 respectively. PCF1 and PCF2 are connected to CDF1 and CDF3 respectively through the addresses and port numbers. Redis is deployed in CDF, and the sentinel mode of Redis is used to monitor the status of the main CDF.

[0017] Furthermore, in step 1.2, the same network security policies are configured, including: quality of service policies to determine the priorities and bandwidth limits of different types of traffic to ensure service performance; mobility management policies to manage user switching policies between different access networks or cells to ensure service continuity and smoothness; access policies to control user access rights and conditions to the network, such as access control based on user identity or subscription information.

[0018] Furthermore, in step 2, after PCF1 and PCF2 are powered on and run normally and enter the working state, they actively register with NRF through Nnrf_NFManagement_NFRegister, and send heartbeat messages to NRF regularly through Nnrf_NFManagement_NFUpdate according to the configuration cycle. NRF uses the heartbeat message to detect and maintain the working state of PCF, specifically including:

[0019] Step 2.1: Under normal circumstances, the UE initiates a session establishment request, and the SMF requests the PCF address from the NRF. At this time, the NRF returns the network element information of PCF1 to the SMF.

[0020] Step 2.2, SMF sends a policy association establishment request (Policy Association Establishment Request) to PCF, the request contains the user identification (Supi) of the user equipment, Pdu session information, and Dnn (data network name) information;

[0021] Step 2.3: After receiving the request from SMF, PCF makes a policy decision based on the request content and the user's subscription information. The policy decision involves QoS parameters, traffic filtering rules, and billing rules.

[0022] Step 2.4: PCF generates policy control information related to the session based on the request and policy rules, and saves it locally. The user information saved here includes: QoS requirements, charging rules, and priority of the user session. PCF returns the generated policy information to SMF, and the user goes online successfully.

[0023] Step 2.5, after the user successfully goes online, PCF1 determines whether CDF1 is available. If CDF1 works normally, the user's online information IMSI, user billing information, network location information, policy information and location information are saved in CDF1. After successful saving, the database replication function is needed to synchronize the user information to CDF2. The master-slave replication of redis is used to send the operation records in CDF1 to CDF2. CDF2 synchronizes data according to the operation records. If CDF1 does not work properly, CDF2 is connected to save the user's online information to CDF2.

[0024] Step 2.6: After the synchronization between CDF1 and CDF2 is completed, PCF1 synchronizes the user data to DC, and then DC synchronizes the data to CDF3 and CDF4;

[0025] Step 2.7, if a policy update occurs, after receiving the policy update request from SMF, PCF re-evaluates the current policy control based on the user's subscription information, session status and predefined policy rules. After completing the policy decision, PCF will generate new policy control information, including QoS parameters and charging rules;

[0026] Step 2.8, update the newly generated policy information to CDF1, CDF1 sends the data operation record to CDF2, synchronously updates the user information in CDF2, and repeats step 2.6;

[0027] Step 2.9, PCF sends a policy update response (Policy Update Response) to SMF, and the policy update response contains new policy control information, including: QoS parameter adjustment, updated charging rules;

[0028] Step 2.10, when the session release process is triggered, SMF sends a Policy Association Termination Request to PCF, and the request contains the identification information of the session, including the PDU session ID and the user identification (SUPI);

[0029] Step 2.11: After receiving the session release request from SMF, PCF searches for the policy control information of the QoS parameters and charging rules related to the session according to the request content, and deletes all policy control information related to the session, including the rules and decisions generated during session establishment or policy update;

[0030] Step 2.12: After deleting the policy information, the cache related to the session needs to be cleaned up to ensure that system resources are released. PCF requests CDF1 to clear the relevant data of the corresponding session, including: session identification information, policy control rules (qos, network slicing rules), session management information (session status, session duration, charging rules, etc.). After the deletion in CDF1 is successful, CDF1 sends the operation log to CDF2. CDF2 synchronizes the deletion and update according to the operation log, and repeats step 2.6.

[0031] Step 2.13, PCF sends a policy association release response to SMF, notifying SMF that the session policy has been released successfully.

[0032] Furthermore, in step 3, under abnormal circumstances, when CDF1 fails and goes down, the state of CDF1 is automatically monitored using the redis sentinel mode, specifically including:

[0033] Step 3.1: When a CDF1 fault is detected, CDF2 is automatically upgraded to the primary one;

[0034] Step 3.2: If PCF1 fails and goes offline, the heartbeat link between PCF1 and NRF is broken, NRF determines that PCF1 is unavailable, and SMF cannot communicate with PCF1 normally. It then requests NRF for a usable PCF again. NRF sends the registration information of PCF2 to SMF, and the traffic carried by PCF1 is taken over by PCF2. PCF2 obtains the context status data from CDF3 and continues to process the service.

[0035] Furthermore, in step 3.2, if PCF1 goes offline due to a fault:

[0036] Step 3.21: CDF1 fails and the communication between PCF1 and CDF1 is interrupted;

[0037] Step 3.22: Sentinel detects CDF1 failure;

[0038] Step 3: Upgrade CDF2 to the main CDF. ;

[0039] Step 4: PCF1 and CDF2 resume communication.

[0040] Furthermore, when PCF1 fails, the PCF processing steps are as follows:

[0041] Step 1: PCF1 fails and the heartbeat between NRF and PCF1 is interrupted.

[0042] Step 2: The communication between SMF and PCF1 is interrupted.

[0043] Step 3: SMF requests the PCF address from NRF again.

[0044] Step 4: NRF responds to SMF with the registration information of PCF2.

[0045] Step 5: SMF and PCF2 resume communication.

[0046] Step 6: PCF2 obtains user information from CDF3.

[0047] Compared with the related prior art in this technical field, the PCF remote disaster recovery method of the present invention has the following superior technical effects:

[0048] 1. The PCF off-site disaster recovery method of the present invention deploys PCF in multiple regions, so that even if a failure occurs in one region, the system can continue to operate, thereby ensuring the high availability of the system.

[0049] 2. The PCF off-site disaster recovery method described in the present invention provides data backup and recovery functions for PCF off-site disaster recovery, can back up data in real time, reduce the risk of data loss, and ensure data integrity and reliability.

[0050] 3. The PCF off-site disaster recovery method described in the present invention can quickly restore business when a failure occurs, reduce downtime and recovery time, and the backup system can respond quickly to ensure business continuity, which plays an important role in ensuring network stability and reliability and reducing operational risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is the PCF remote disaster recovery method of the present invention, and a schematic diagram of the PCF processing method under normal circumstances;

[0052] Figure 2 It is a method for PCF remote disaster recovery of the present invention, and a schematic diagram of a method for handling a CDF failure;

[0053] Figure 3 It is a schematic diagram of the PCF remote disaster recovery method of the present invention and a processing method when a PCF failure occurs. DETAILED DESCRIPTION

[0054] The specific implementation of the PCF remote disaster recovery method of the present invention is described in detail below with reference to the accompanying drawings.

[0055] The drawings provided for the method described in the present invention are used to provide further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their description are used to explain the present invention and do not constitute improper limitations on the present invention.

[0056] Example

[0057] like Figure 1As shown, in a normal scenario, the PCF processing steps of the method of the present invention are as follows:

[0058] Step 1: SMF requests the PCF address from NRF.

[0059] Step 2: NRF responds the registration information of PCF1 to SMF.

[0060] Step 3: SMF sends a policy association establishment request to PCF.

[0061] Step 4: PCF makes a policy decision and generates policy control information and charging rules applicable to the session.

[0062] Step 5: PCF returns the policy control and charging rules to SMF in the form of a policy association establishment response.

[0063] Step 6: After the user successfully goes online, the user information is saved in CDF1.

[0064] Step 7: CDF1 synchronizes the data to CDF2.

[0065] Step 8: CDF1 and CDF2 synchronize the data to DC.

[0066] Step 9, DC synchronizes user data to CDF3 and CDF4;

[0067] Example 2

[0068] like Figure 2 As shown, when CDF1 fails, the PCF processing steps of the method of the present invention are as follows:

[0069] Step 1: CDF1 fails and the communication between PCF1 and CDF1 is interrupted.

[0070] Step 2: Sentinel detects CDF1 failure.

[0071] Step 3: Upgrade CDF2 to the main CDF.

[0072] Step 4: PCF1 and CDF2 resume communication.

[0073] Example 3

[0074] like Figure 3 As shown, in the method of the present invention, when PCF1 fails, the PCF processing steps are as follows:

[0075] Step 1: PCF1 fails and the heartbeat between NRF and PCF1 is interrupted.

[0076] Step 2: The communication between SMF and PCF1 is interrupted.

[0077] Step 3: SMF requests the PCF address from NRF again.

[0078] Step 4: NRF responds to SMF with the registration information of PCF2.

[0079] Step 5: SMF and PCF2 resume communication.

[0080] Step 6: PCF2 obtains user information from CDF3.

[0081] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with the technical field within the scope disclosed in the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A PCF remote disaster recovery method, characterized in that: The following steps are involved: Step 1: Deployment of multi-region PCF environment; Step 2: After PCF1 and PCF2 are powered on and enter the working state, they actively register with NRF through Nnrf_NFManagement_NFRegister, and send heartbeat messages to NRF regularly through Nnrf_NFManagement_NFUpdate according to the configuration cycle. NRF uses heartbeat messages to detect and maintain the working state of PCF. Step 3: In abnormal situations, when CDF1 fails, use the Redis sentinel mode to automatically monitor the status of CDF1; Step 4, PCF2 determines whether CDF3 is available. If CDF3 works normally, the user information is obtained from CDF3 and saved locally in PCF. If CDF3 does not work normally, the user information is connected to CDF4 to obtain the user information from CDF4.

2. According to the PCF remote disaster recovery method of claim 1, in step 1, the multi-region PCF environment deployment specifically includes: Step 1.1, deploy a data center (DC); Step 1.2: Select appropriate geographical locations to deploy a multi-region PCF environment based on business needs, and ensure that there is a stable network connection between regions to support data synchronization and failover. Deploy PCF1 and PCF2 in the two selected regions respectively, ensure that the regional environments can communicate with each other, and configure the same network security policies. Step 1.3, build CDF1 and CDF2 environment, CDF1 and CDF2 network elements are mutually active and standby, CDF1 is responsible for storing user online information in PCF1 network element; Step 1.4, build CDF3 and CDF4 environment, CDF3 and CDF4 network elements are mutually active and standby, CDF3 is responsible for storing user online information in PCF2 network element; Step 1.5: All CDFs are linked to DC. When CDF is used as the primary, it synchronizes data to DC. When CDF is used as the backup, it needs to synchronize user information from DC. Bidirectional synchronization is used between CDF and DC. Step 1.6, configure the communication addresses and port numbers of CDF1, CDF2, CDF3, and CDF4 respectively. PCF1 and PCF2 are connected to CDF1 and CDF3 respectively through the addresses and port numbers. Redis is deployed in CDF, and the sentinel mode of Redis is used to monitor the status of the main CDF.

3. According to the PCF remote disaster recovery method of claim 1, in step 1.2, configuring the same network security policy specifically includes: Quality of service policies to determine the priorities and bandwidth limits for different types of traffic to ensure service performance; Mobility management strategy, which manages the switching strategy of users between different access networks or cells to ensure the continuity and smoothness of services; Access policy controls the rights and conditions for users to access the network.

4. According to the PCF remote disaster recovery method of claim 1, in step 2, after PCF1 and PCF2 are powered on and run normally and enter the working state, they actively register with NRF through Nnrf_NFManagement_NFRegister, and send heartbeat messages to NRF regularly through Nnrf_NFManagement_NFUpdate according to the configuration cycle. NRF uses the heartbeat message to detect and maintain the working state of PCF, which specifically includes: Step 2.1: Under normal circumstances, the UE initiates a session establishment request, and the SMF requests the PCF address from the NRF. At this time, the NRF returns the network element information of PCF1 to the SMF. Step 2.2, SMF sends a policy association establishment request to PCF, the request contains the user ID of the user device, Pdu session information, and Dnn information; Step 2.3: After receiving the request from SMF, PCF makes a policy decision based on the request content and the user's subscription information. The policy decision involves QoS parameters, traffic filtering rules, and billing rules. Step 2.4: PCF generates policy control information related to the session based on the request and policy rules, and saves it locally. The user information saved here includes: QoS requirements, charging rules, and priority of the user session. PCF returns the generated policy information to SMF, and the user goes online successfully. Step 2.5, after the user successfully goes online, PCF1 determines whether CDF1 is available. If CDF1 works normally, the user's online information IMSI, user billing information, network location information, policy information and location information are saved in CDF1. After successful saving, the database replication function is needed to synchronize the user information to CDF2. The master-slave replication of redis is used to send the operation records in CDF1 to CDF2. CDF2 synchronizes data according to the operation records. If CDF1 does not work properly, CDF2 is connected to save the user's online information to CDF2. Step 2.6: After the synchronization between CDF1 and CDF2 is completed, PCF1 synchronizes the user data to DC, and then DC synchronizes the data to CDF3 and CDF4; Step 2.7, if a policy update occurs, after receiving the policy update request from SMF, PCF re-evaluates the current policy control based on the user's subscription information, session status and predefined policy rules. After completing the policy decision, PCF will generate new policy control information, including QoS parameters and charging rules; Step 2.8, update the newly generated policy information to CDF1, CDF1 sends the data operation record to CDF2, synchronously updates the user information in CDF2, and repeats step 2.6; Step 2.9: PCF sends a policy update response to SMF. The policy update response includes new policy control information, including: QoS parameter adjustment and updated charging rules. Step 2.10, when the session release process is triggered, SMF sends a policy association release request to PCF, and the request contains the identification information of the session, including the PDU session ID and the user ID; Step 2.11: After receiving the session release request from SMF, PCF searches for the policy control information of the QoS parameters and charging rules related to the session according to the request content, and deletes all policy control information related to the session, including the rules and decisions generated during session establishment or policy update; Step 2.12: After deleting the policy information, the cache related to the session needs to be cleaned up to ensure that system resources are released. PCF requests CDF1 to clear the relevant data of the corresponding session, including: session identification information, policy control rules, and session management information. After the deletion in CDF1 is successful, CDF1 sends the operation log to CDF2. CDF2 synchronizes the deletion and update according to the operation log, and repeats step 2.

6. Step 2.13, PCF sends a policy association release response to SMF, notifying SMF that the session policy has been released successfully.

5. According to the PCF remote disaster recovery method of claim 1, in step 3, under abnormal circumstances, when CDF1 fails and goes down, the state of CDF1 is automatically monitored using the redis sentinel mode, specifically including: Step 3.1: When a CDF1 fault is detected, CDF2 is automatically upgraded to the primary one; Step 3.2: If PCF1 fails and goes offline, the heartbeat link between PCF1 and NRF is broken, NRF determines that PCF1 is unavailable, and SMF cannot communicate with PCF1 normally. It then requests NRF for a usable PCF again. NRF sends the registration information of PCF2 to SMF, and the traffic carried by PCF1 is taken over by PCF2. PCF2 obtains the context status data from CDF3 and continues to process the service.

6. According to the PCF remote disaster recovery method of claim 5, in step 3.2, if CDF1 goes offline due to a fault: Step 3.21: The communication between CDF1 and CDF1 is interrupted; Step 3.22: Sentinel detects CDF1 failure; Step 3.23: Upgrade CDF2 to the main CDF; Step 3.24: PCF1 and CDF2 resume communication.

7. According to the PCF remote disaster recovery method of claim 5, in step 3.2, when PCF1 fails, the PCF processing steps are as follows: Step 1: PCF1 fails, and the heartbeat between NRF and PCF1 is interrupted; Step 2: The communication between SMF and PCF1 is interrupted; Step 3: SMF requests the PCF address from NRF again; Step 4: NRF responds to SMF with the registration information of PCF2; Step 5: SMF and PCF2 resume communication; Step 6: PCF2 obtains user information from CDF3.

Images