Wireless sensor networks offer many beneficial applications in various fields. However, because sensor devices are limited in their resources and susceptible to a variety of novel attacks, even a few malicious adversaries can easily spread deceitful data and make the networks be in confusion without great efforts. Therefore, it is essential to distinguish forged data of illegal nodes from innocent data of legal nodes in sensor networks. In this paper, to make resilient wireless sensor networks, we propose a trust evaluation model which can identify trustworthiness of sensor nodes in order to filter out malicious nodes' deceitful data