With the rapid development of Internet of Things (IoT) technology, the widespread deployment and heterogeneity of IoT devices bring new challenges to device identification. For network administrators, understanding the IoT devices connected to the network is very important. However, there are still some issues with existing IoT device identification methods, including the dependence on a large number of features, vulnerable features (such as IP addresses), and payloads. In response to these issues, an IoT device identification method based on network traffic is proposed in this paper. The method utilizes network traffic data of IoT devices over a short period of time, extracts protocol statistical features and flow-level statistical features, and thus achieves the identification of IoT device categories and device types. The proposed method was evaluated on two publicly available datasets, and the results demonstrate that the proposed method has superiority and can accurately and effectively identify devices.