In order to design a low-calculation and high-precision TOA (time of arrival) estimation algorithm for UWB (ultra wideband) based wireless sensor network (WSN), a two-step TOA estimation method which jointly employs energy-detection (ED) and match-filtering (MF) is proposed in this paper. Based on analyzing the principles of the two-step method, it is pointed out that the success rate of DP (direct path) block detection in the first step and the setting of MF-threshold-factor in the second step are the key issues that affect the performance of the method. Algorithm selection of the first step and setting of the energy integration interval, which are the two factors that affect the success rate of DP block detection, are discussed through simulations. The idea of DMR (DP to minimum energy sample ratio) based MF-threshold-factor selection is proposed, and the mathematical model of the relationship between DMRs and the optimal MF-threshold-factors are built. Results show that the proposed two-step method greatly outperforms the one-step energy-detection based non-coherent method, while largely decreases the computational complexity compared to the one-step match-filtering based coherent method, so that the two-step method is more appropriate for application to sensor nodes which need to be designed with low complexity and low power consumption.