The credit allocation of prepaid services in mobile telephone networks
From 2G to 3G, prepaid service has been one of the most important services in mobile communication networks. As the systems evolve, prepaid services are also extended from providing only voice
communication to cover voice, data and multimedia services.
Today, a prepaid user can use multiple services simultaneously. In addition, multiple users can share one account. It is now a great challenge for the system to achieve real-time billing and to minimize the bad debt.
In 3GPP specs, there is no description about how to allocate the user credit to the services. If the mobile system allocates insufficient credit to a prepaid service, the prepaid service control point (P-SCP) will be busy in checking the user’s credit. The number of
communication between the P-SCP and the network elements will increase, which consumes lots of system resources. On the other hand, if the amount of credit is excessively allocated, the other services may be restricted to use at the same time. Besides, when the credit becomes lower and lower, prepaid services may be forced to terminate due to insufficient credit. The operator will be flushed with lots of complaint. Hence, the credit allocation is a key to the success of the mobile prepaid service.
We will use F-ARIMA model developed by Kettani to predict the amount of credit which is needed by a prepaid service. The P-SCP will take this value as a reference when it allocates the prepaid credit. Compared with the traditional method – Wavelet, the advantages of F-ARIMA include faster computation and smaller confidence intervals of the parameters.
We assume that a prepaid user can use multiple services
simultaneously in our system such as vice, email and packet data transfer services. We will compare the signaling cost of F-ARIMA and other methods when the service times are exponentially, Erlang, Gamma or Pareto distributed. In addition, we will integrate F-ARIMA with the reserve credit method and classify all services. When the user credit is below a threshold, we compare their performance to find out the minimal forced-termination probability.
At last, we use the freeware “OpenDiameter” to implement a credit
control application server. We use the server to simulate the function of a P-SCP to accept a service request, compute the allocated credit and assign this credit to the prepaid service.
Keywords: Credit Allocation, F-ARIMA, Prepaid Service, Prepaid Service Control Point
