LAUSR

In this work, a newly dynamic step size-based 5G charging mechanism was proposed with the introduction of random forest algorithm, user daily traffic forecast model, and RG distribution model for the pressure release on charging system caused by the growing consumption demands based on the behavior feature selection of subscribers. After the verification of the dynamic step size mechanism, optimization analysis was performed from three aspects, including authorization step statistical results, Charging Data Record, and timely reminders by utilizing the real monthly traffic usage data of 13,000 users from China Telecom. Result shows that the amount of charging messages between the network elements and the charging systems decreases by about 35%, and the timeliness rate of reminder service increases by approximately 30%. This research realizes an innovative updating of traditional charging systems to effectively improve the efficiency of the 5G charging system and achieves a reduction in system maintenance and resource occupation. Along with the improvement of service quality and user experience, this work provides an innovation method for operators and different associated businesses to tackle new challenges in 5G.