In order to accurately provide the extent of water deficit and thus to offer a scientific reference for the precision irrigation, we explored the application of wireless sensor networks to precision irrigation system based on the acoustic emission principle for crop water stress. Adaptive weighted data fusion algorithm was adopted to improve the precision of acoustic emission signal and cluster-based multi-hop routing algorithm was proposed to reduce energy consumption of node transmitting data. By a NB100 gateway, the system realized bridging between wireless network and wired network. It runs distributed and possesses many advantages, such as good robustness, extensibility and scalability. Simulation experiments showed that networking was correct and the energy consumption of wireless transmission was above 60 percent in total energy. The system enables user to precisely acquire the crop water requirement information and thus to implement the precision irrigation remotely, and can be effectively applied to some water-saving agriculture fields, for example, the cropland, the nursery garden, the greenhouse, etc.