Abstract: ob scheduling system is the core of data mule and challenging problem in wireless sensor network. How to improve the efficiency of resource and the maximum profit are the ultimate goal of service providers. In at first, by establishing the preemptive priority queuing modle of G/G/l,to meet the QoS of the transmit data at different levels,the linear piecewise functions are used to make the system cost minimum. Furthermore, based on the cost function and profit function of queuing system, to get dynamic allocation rate, and to get the corresponding value which service users and service providers are satisfied with. The results show that, the dynamic resource scheduling system, not only can ensure the different data's QoS requirements, but also can provide the maximize profits for the wireless sensor network service providers.

Key words: Job scheduling system, Data mule, Preemptive priority, Profit