In view of the security threats in the coal mine environmental monitoring system, this paper proposes a method for distributed hierarchical information fusion based on improved Kalman filter and SVM. The sensors in the system are divided into source nodes and sink nodes. In the source, the Kalman filtering are used, and in the sink SVM is used. The algorithm can reduce the energy consumption and data collision. Simulation result verifies the reliability of this algorithm, and it improves the fusioning precision of the wireless sensor network .