Abstract. In this technical survey, we introduce the state-of-the-art signal-to-noise rate (SNR) estimation techniques for direct sequence ultra-wide bandwidth (DS-UWB) system. The SNR estimation is important to improve the system performance in a multiuser wireless sensor network (WSN). We demonstrate the effectiveness of using regression analysis of features extracted from code mapping to derive a new SNR estimator. And compared the result with another estimator derived by long short-time memory (LSTM) recurrent neural networks (RNN) in terms of their root-mean-squared error. Numerical result show that with a proper length of sequence the proposed method can reach an excellent performance when operating in a DS-UWB system.