Frequency Estimation of Sinusoid Based on the Dual-Threshold Decision Modified Rife Algorithm

In order to make a good tradeoff between the complexity and performance for frequency estimation of sinusoid, a frequency estimation algorithm with high estimation accuracy and low computation was proposed based on the dual-threshold decision fixed Rife algorithm. The frequency was divided into three domains with different processing schemes, which would be suitable for low power consumption and low-delay applications. Simulation results show that the method overcomes the inherent flaw of the classic Rife algorithm, and the RMS (root mean square) error for the estimated frequency is only 0.5 dB from the CRB (Cramer-Rao Bound). Compared with other improved Rife algorithms, the new approach can significantly reduce the required computation under the premise of guaranteeing the performance of estimation, be propitious to the engineering.