Explore key technologies for building multi-network integrated user data centers

With the development of IT software and hardware technology and business development, more and more communication networks are introduced and become standards in addition to the traditional PSTN network and 2G network, such as 3G network, LTE network, IMS network, WiMAX network, and Softswitch Network. This leads to a large increase in user data.

With the surge in user data in the network, the original methods of storing and managing user data in their respective networks are increasingly unable to adapt to the development of the network. Scattered Data Management leads to high OPEX, difficulty in eliminating junk data, difficulty in ensuring data consistency, difficulty in ensuring data security, and lack of effective means for data mining and analysis. The external interfaces of data storage network elements are different, as a result, data access interfaces are complex, interfaces are tightly coupled with businesses, and data cannot be transparently transmitted. As a result, new businesses cannot be quickly and effectively carried out, and user data management problems have become increasingly apparent.

To this end, the industry has proposed the concept of a unified user data model, which combines the data of the same user distributed on different network elements, sets the user ID as the basic identifier, and organizes the data according to the unified data structure, as the only user data source for all networks. In terms of physical implementation, the unified user data model is stored in the unique network data storage network element: the central database (CDB), which stores the same user data model, this service ensures data consistency and reliability to ensure security. It provides open interfaces unrelated to services and data, and provides data access services for other data query network elements.

Multi-network integration terminal data, access/core network data, business data, and Internet business data. applications and data are decoupled for unified management. A unified user data center is established to simplify the network, reduce OPEX/CAPEX, improve data security, shorten new business release time, promote business innovation, and lay the foundation for operators to provide competitive business integration.

2. Converged data center architecture

Converged data centers must have the following features:

Coupling: The business processing logic and data storage are loosely coupled. Data of different networks can be stored in the same integrated center database. data modification of one network does not affect data of the other network.

Open data access interfaces: the database must provide open interfaces for data to access data using network elements.

Distributed Storage: A Converged data center is composed of multiple nodes. The data of each node is fully synchronized. These nodes can be geographically distributed anywhere in the IP network to provide high reliability. To meet the features of the above databases, this article proposes a central database architecture as shown in 1:

Figure 1 database architecture of the converged Center

The integration center database consists of two layers: the application layer and the database layer. The Application Layer completes the processing of various business application logic, such as HLR, AAA, SCP, and other network elements. I do not need to perceive the data organization and storage methods, and the business and data are loosely coupled. The database layer stores and manages data.

The Ud interface is an open interface between the application layer and the database layer. For example, the LDAP interface provides data access. Its interfaces have nothing to do with the specific data structure, and meet the requirements of converged center database services and loose data coupling and open data access interfaces. Because open interfaces are used between the application layer and the database layer. Therefore, these two-layer devices can be composed of devices from different manufacturers. This database architecture provides excellent networking flexibility.

The application layer and database layer are composed of multiple nodes that can be geographically distributed anywhere in the IP network. The data of each node in the database layer is completely synchronized. Therefore, this distributed architecture has a natural disaster recovery function, making the entire system highly reliable.

The database layer consists of two parts: resource management and data management. The Resource Management module includes physical device management and data storage management modules to provide resource support for data storage.

Data management includes:

(1) data conversion/provision: Provides interfaces for the application layer to access data, accepts data access requests from the application layer, and returns Data Access results to the application layer;

(2) Access Control: access control is implemented at the application layer of the eight data layers to prevent system attacks;

(3) Data View Control: for the sake of system security, a custom view is provided for each application. Different applications should access different views;

(4) Data Consistency Control: Control Data Synchronization and consistency between the master and slave nodes to ensure data integrity;

(5) data backup/recovery: Controls data backup and node recovery and data recovery when new nodes are added;

(6) overload control: when the actual load exceeds the designed load of the system, the overload control function will take effect and the transaction requests that cannot be processed due to system capacity restrictions will be limited, to protect the normal operation of the system;

(7) policy control: provides different data access policies and data access priorities for different application types or different application instances.

3. Unified User Data Model Modeling

To integrate and manage user data in multiple network elements, you must solve the problem of how to organize user data. This article discusses a data organization method of the directory Information Tree (DIT, Di rectory Information Tree). 2 shows that it can efficiently manage user data.

Figure 2 directory information tree data organization model

User Data is organized in the form of DIT. "ROOT" is the ROOT node of the tree. All access to data starts from the "ROOT" node. Under the "ROOT" node, different leaf nodes are organized by users. For example, all data about user 1 is organized under the "Subsc riberl" node. Different business data of each user is organized and placed under different leaf nodes. For example, under the "Prepaid Service" node, the data is related to the user's "Prepaid Service. Generally, any user-based data can be loaded into the DIT as a subtree.

In the DIT model, each user's data is identified by a unique UID (uniied Identification). If other user IDs are used to access user data, external network elements are mapped to UID. then access user data. The ing between other user IDs and UJD is stored in the "ROOT" node. For example, the user data is accessed using MSISDN. First, MSISDN is mapped to a UID on the "ROOT" node, and then the user data is queried as the UID. That is, different service data of a user is stored as a sub-tree using the same user ID (UID). external network elements can access user data by using IMSI and MSISDN with different identities. A higher-level node is called a "parent" node, and a lower-level node is called a "child" node. A data access path is formed between a "child" node and a "parent" node. Therefore, after the data model is created, the data access path is determined accordingly.

The leaf node of the directory information tree defines the data type of its node through the object class objectClass. The specific data is an instance of the object class, and an instance is composed of different attributes.

By establishing a unified user data model and organizing data based on users, You can organically organize user data originally distributed across multiple network elements under a "subtree, the unified UID can be used to access user data that is originally scattered across multiple network elements of the same user, thus realizing the integration and unified management of user data.

4. Data Access Process

The Application Layer accesses data in the database layer through an open data access interface (such as the LDAP Interface), as shown in 3.

The data access process is described as follows:

(1) When an external network element needs to establish communications with the integrated data center during business processing, the external network element initiates a business request message to the application layer of the integrated data center:

(2) After the application layer of the integrated data center receives a business request, it executes the corresponding business application logic. In this process, the application layer sends a data access request message to the database layer to access user data;

(3) After receiving the data access request message, the database layer executes the access control at the application layer, verifies the access permissions at the application layer, and checks the access priority at the application layer;

(4) After the access control check is passed, the database layer obtains the data to be accessed at the application layer based on the data view at the application layer;

(5) The database layer returns a data access Response Message to the application layer, carrying the data access result (successful or failed );

(6) After the application layer receives the data access result, it continues to execute the business application logic and interact with the external network element to complete the business processing requested by the external network element.

Figure 3 data access process