A Proposal for the Management of Mobile Network's Quality of Service (QoS) using Data Mining Methods

Today, the challenge for the service operators is not only to attract and subscribe new users but to retain already subscribed users. To gain a competitive edge over other service operators, the operating personnel have to measure the services provided to their users and the network performance in terms of Quality of Service (QoS) at regular periods. By analyzing the information in these measurements, they can manage the quality of service, which helps to improve their service and network performance. But due to the heavy increase in the number of users in recent years, they find it difficult to elicit essential information from such a large and complex data to manage the QoS using the existing methods. It is here that the recently developed and more powerful data mining methods come in handy. In this paper we proposed how data mining methods can be used to manage the mobile network QoS. We describe three data mining methods: Rough Set Theory, Classification and Regression Tree (CART), and Self Organizing Map (SOM).