Frequency estimation by 1-bit quantization and table look-up processing

A method for fast frequency estimation by table lookup processing (FFETL) is presented. The estimation is based on data that has been quantized at one bit per sample, and all data processing is represented by a single table look-up operation, resulting in O(l)-complexity. The performance of the new method is compared with the proper Cramér-Rao bound for one bit quantized data, and the MLE for unquantized data by aid of Monte Carlo simulations. FFETL is shown to be (almost) statistically efficient over a wide range of SNR, as encountered in practical applications. Practical aspects such as implementational issues, and performance limitations due to quantization are discussed in some detail.