| 摘要 |
A method is disclosed for balancing a load of clients of a network across a plurality of communications providers. Connection data that includes information on attempts to connect to the network and whether each attempt was successful is collected. A time-independent demand curve for each of the communications providers representing a likelihood of successful connection as against client load is determined, based on the collected connection data. The load of clients is then apportioned across the communications providers based on the time-independent demand curve for each of the communications providers.
|
