The position function is becoming more and more important for applications in WSNs, so a novel localization model, which is based fuzzy, is aimed at the localization problem. In this model, a space is divided into several small areas by some swatch nodes; every swatch node has its unique signal vector. An unknown node’s position can be calculated by computing the approach degree between the signals of the swatch and unknown nodes. This model adopts the RF signal to locate the unknown node. It cannot only avoid the overlapped error in a range-based localization model, but can also avoid the high complex. The experimental results in NS-2 show that this model performs well and is suitable for applications in WSNs.