Charging users for a server operator billing involves performing a number of operations. Each service provided by the operator or service provider customers , charged at a separate scheme , as the bottleneck in the process of providing services are bilingual system. Operation performed is limited. In case of exceeding its application is considered lost and the user is notified of the inability to obtain services. In this situation, the operator has monetary losses. In systematic failures in service deteriorates reputation . Therefore, it is important to make a control input streams so as to reduce the possibility of exceeding the service time of the user's requests. The process of managing incoming applications to the server should consider charging the features of charging large companies:
• charging process involves a sequence of standard operations , speed and resources for the implementation of which depends on the type of service that serviced
• a variety of services that are billed differently ,
• extremely high number of requests for billing ,
• irregularity of the input stream , applications received according to Poisson.
Scheme smoothing the input load is a set of values of the maximum number of requests (the sequence {ki}), arriving at the input of the system in a short time interval in a given sequence. Number of elements in a sequence of n is chosen so that the equation was carried out , where t - the average time of the system. It’s necessary to choose a sequence {ki} to satisfy conditions:
1. Applications that are both served by the system must use the volume V of resources close to the total maximum number of resource Vmax. Dispersion sequences such amounts shall be minimized.
2. Dispersion elements of the sequence {ki} should be minimal.
The method of solution of choice sequence {ki} by using genetic algorithm. When using the proposed scheme smoothing the input load is provided by the maximum allowable flow uniformity requests. Since overlooked the sequence of operations performed on the application server to the mobile operator, the amount of resource that is required for these operations. Selected sequence of maximum permissible values of number of applications coming into the system in a short period of time.