Abstract: Aiming at the deficiency in controlling arrears risk of traditional balance allocate algorithm, in Online Charging System(OCS), based on customer’s consumption history, using cluster analysis method, this paper designs an adaptive balance allocate algorithm based on user habit, which can be optimized by table looking-up and fractionation strategy. Test results show that the optimized algorithm can highly reduce the arrears risk, improve user experience and system performance of OCS.

Key words: Online Charging System(OCS), real-time charging, cluster analysis, balance allocation

摘要： 针对传统余额分配算法存在不能有效控制欠费风险的问题，在在线计费系统(OCS)中，基于用户的消费历史记录，采用聚类分析法，设计一种基于用户习惯的自适应余额分配算法，利用分段与查表反算策略对其进行优化。测试结果表明，优化算法能有效降低欠费风险，提升用户体验和OCS系统性能。

关键词: 在线计费系统, 实时计费, 聚类分析, 余额分配

CLC Number: 

• TP311