Today’s radio access technologies demand flexible, high performance and power-efficient computing. Software-defined radios offer a high degree of flexibility but their performance is limited in energy-constrained environments. Reconfigurable computers exhibit a good trade-off between performance and power consumption. The plethora of reconfigurable devices and architectures suggests using middleware with virtualization and resource management capabilities. Cognitive radios can take benefit of such a middleware approach for obtaining waveform resource demands and employing advanced resource management techniques. This paper compares two approaches for dynamic partial reconfiguration and their suitability for middleware. Our reconfiguration approaches are novel in the sense that they use an intermediate memory as a buffer to isolate the bitstream download from the actual reconfiguration. Performance and resource overhead is measured and analyzed for both approaches.