The Wireless Sensor Networks (WSNs) have been deployed in many applications of daily life. Sensors in networks can be located in extreme environments or even the replacement is impossible. Therefore, it is necessary to design the high-quality transmission system. The fault management plays a key role in this system. Many studies in this area have approached the difference methods for fault diagnosis. Machine learning algorithms also have been deployed to manage and detect fault in WSNs. In this paper, we propose the fault detection for WSNs using the Support Machine Learning (SVM) algorithm.