one of the characteriristics of wireless sensor network is to produce high-speed data streams. the data stream, which is continuously generated at a rapid rate, differs fom traditional static data sets. thus, ordinary data mining techniques are not suitable to process this kind of data. this is because the ordinary data mining techniques are designed to handle static data sets where severalpasses over the stored data are possible. this condition is different with data streams where only a single pass processing method for each data is allowed. at the some time, wireless sensor devices are constrained by limited resource availability. thus, energy efficiency and good resource management are vital for on-board devices processing techniques. this research would like to measure the effecctiveness of resource-aware framework for wireless sensor network devices. we evaluated resource adaptive online clustering and calrification currently have ben developed on the novel sun microsystemTM Small Programmable Object Technology (Sun SPOT) platform. experimental result show that resource-aware framework works well against scarce resources and are able to improve significantly in resource utilization while maintaining acceptable accuracy level.