Abstract: Most of the current WSN research platforms are based on Mica nodes and TinyOS. However, they're difficult to integrate with the existing mature source codes and difficult to port to other platforms. These micro-controller based solutions are also quite weak to fulfill high performance tasks such as image processing and PAN coordination. Furthermore, the new invented nesC language also holdback its acceptance among industry application developers. Considering the above shortcomings, an ARM7 based powerful WSN research platform, named OpenWSN, is designed and implemented. OpenWSN adopts bus architecture to connect different modules in hardware design and uses layered OOD/OOP in software design. The component-based and service-based architecture makes it easy to integrate with current mature resources. The node has 32KB SRAM and up to 512KB flash memory. It can run between 11-60MHz and using DMA to move data inside memory. These features enable OpenWSN run much faster than general Mica or Telos nodes. The power consumption is a bit higher than Mica's Atmega128L solution because ARM7 processor consumes more energy than Atmega128L micro processor. However, due to its high performance, OpenWSN node can finish more tasks and spend more time in sleep status. This in turn decreases the power consumption. The experiment verifies the solution and confirms OpenWSN's effectiveness in high performance applications such as vibration measurements.