With the development of wireless networks, the amount of multiple services increased sharply in recent years. High quality multiple services with low price are urgently needed especially in new generation mobile communication systems, e.g., 3G/LTE networks. It is important to enhance the availability of data service resources. Services have strong association which are used by clients with similar behavior habits in networks. Such feature results in service behavior convergence （SBC） and its utilization will enhance resource efficiency. This paper proposes two applications of service behavior： service prediction and a scheduling algorithm which enhances bandwidth efficiency. Convergence cells are classified according to SBC and hot-spot services are broadcasted separately in each convergence cell. It is demonstrated by stimulation that the bandwidth is saved 80% more than classical cellular system and nearly 20% more than traditional broadcasting system.
