Hardware efficient frequency estimation and tracking using signal autocorrelations

This paper proposes an algorithm for efficient frequency estimation and tracking. The proposed algorithm can be implemented with a small number of multiplications. This is important because in most hardware implementations (e.g. in FPGA) the number of multiplications is the most expensive for silicon area optimization. A correlation based pulse-pair (PP) method, which is an efficient technique for frequency estimation, is used for the initial frequency estimation. It is known that the PP method has a very low number of multiplications, and is ideally suitable for FPGA type hardware implementations, especially at high frequencies. The disadvantage of using the PP method is that the method requires inverse tan operation in the implementation that makes the technique computationally unattractive. In this paper, a new efficient recursive technique for implementing inverse tan operation is proposed. Method for proper frequency estimation and tracking of signals with time varying instantaneous frequencies is also discussed in the paper.