Limited resource allocation and scheduling are important problems in distributed wireless sensor networks (WSNs). Saving energy and real-time performance deserve research for low-cost sensor nodes in the target tracking application. In this paper, we propose a new dynamic-group idea and apply it to our scheduling scheme. An Extended Kalman Filter (EKF) is employed to predict tracking accuracy, and an energy model is adopted to predict the energy consumption. Simulation results show that, compared with the information-driven sensor quering (IDSQ) and the adaptive sensor scheduling strategy, the proposed scheme can achieve significant improvement on real-time property and tracking accuracy for target tracking in WSN. At the same time, it approaches to the adaptive sensor scheduling strategy in energy consumption.