Comparison of accuracy and complexity of advanced frequency estimators

This paper provides performance analysis of advanced frequency estimators that has been recently proposed in the literature for signals in the additive white Gaussian noise. The optimal maximum likelihood (ML) solution is too complicated for practical implementation. Many suboptimal techniques have been presented, which are mainly classified as correlation-based or periodogram-based estimators. The primary aim here is to obtain a fair performance comparison of many recent estimators from both classes under the same simulation environment and for different application scenarios so that to get a better understanding about the differences, most precisely, in terms of accuracy, complexity, frequency acquisition range, signal to noise ratio (SNR) threshold and the sensitivity of these towards different SNR and frequency scenarios.