Method for implementing distributed HLR memory database

Abstract

The invention relates to a method for realizing distributed HLR memory database, wherein it comprises that: a, dividing the data of HLR memory database into different types, defining different memory database servers; b, distributing the database servers on the nodes; c, defining data dictionary for each type of memory database server; d, building memory database; e, server via relative dictionary loads data into memory list; f, building local access interface and remote access interface for the list of memory database server, while the local interface is used to access into the list on local node, and the remote interface is used to access the list on the main server; g, the modification on the memory database list is distributed by the main server to other backup nodes to be synchronously refreshed. The invention can improve the capacity and property of HLR memory database.

Description

technical field [0001] The invention relates to PHS, GSM and 3G mobile communication systems, in particular to a method for implementing user data management by adopting a distributed HLR memory database as a database of a home location register (HLR). Background technique [0002] In PHS, GSM and 3G mobile communication systems, user basic attributes, various subscription data and location information are all stored in their home location register HLR. With the continuous integration of PHS, GSM, CDMA and 3G networks, as a user data center The user data capacity of the HLR grows rapidly, and at the same time, it also puts forward high requirements on the processing capacity of the system. For performance considerations, the current HLR gradually adopts the memory database as the user's real-time data management, and the commercial database as the backup of the user's data storage. [0003] For the sake of simplicity of description, this paper agrees that computer nodes are...

AI Technical Summary

Problems solved by technology

Fig. 1 has described the HLR front-end memory database structure in the WCDMA core network CN; The memory database of the front-end single board adopts simple data to be divided by the number segment, has possessed the data distribution storage function; Under the existing application situation, the single-board support The memory database has great limitations: 1. The memory of the front-end single board is the biggest limitation of the front-end memory database at present. The in-memory database will be limited by the expansion of the number of boards and the processing capacity of the boards; 2. The front-end single-board-level in-memory database has data access efficiency problems. At present, there is a problem of secondary positioning of the foreground database data source in the memory database access. Query the board where the data resides on each board, forward the service request to the board where the data resides, complete the business processing on the board where the data belongs, and then return to the previous board that forwarded the message, and then respond to the business request

Figure 2 shows the HLR background centralized memory database structure in the WCDMA core network CN; the background nodes are configured as high-end servers, and the access of service request messages to the memory database is concentrated in the background server; the advantage of this architecture is that the system is simple to implement, and the background server The memory can be expanded in a large range; but the disadvantage is that the capacity cannot be expanded indefinitely, which cannot satisfy the surge in capacity brought about by network integration and regional data integration. At the same time, a single server is increasingly showing performance bottlenecks

[0006] To sum up, with the integration of user data and the continuous upward movement of network-side data, the current HLR user capacity continues to increase, from the original capacity of hundreds of thousands of users to millions or tens of millions of orders of magnitude. The current front-end memory database will face capacity problem, the background centralized memory database server will face two problems of capacity and performance

Images