Abstract: For a large-scale WSNs deployed in an environment with poor ground transport, it will be better to collect sensor data from WSNs with unmanned aerial vehicle (UAV). However, considering that the UAV has limited resources and there are a large number of sensor nodes in the network, the path planning for UAV is an important factor in whether the sensor data can be collected successfully. This process can be taken as the classical traveling salesman problem (TSP). Considering the large-scale WSNs was deployed homogeneously, an algorithm based on the grid division for the UAV path planning was proposed, which was named as fast path planning with rules (FPPWR). Through the division by a suitable grid, the global path planning can be achieved by merging the path which has been planned according to the primary flight path in the square area. Before merging square path, the paired-operator algorithm will optimize the global path. The experiment indicated that the FPPWR has higher efficiency, besides it also ensures high accuracy.