Efficient time varying frequency tracking using recursive estimation

Abstract

This paper presents performance results of an algorithm for efficient time varying frequency estimation and tracking. The algorithm is implemented with a small number of multiplications, which is important for most hardware implementations. A correlation based Pulse-Pair (PP) method is used for the initial frequency estimation. The PP method requires inverse tan operation that can be implemented using an efficient recursive technique. A simulation example showing proper frequency estimation and compensation for tracking of signals with time varying instantaneous frequencies is also presented in the paper.

Performance analysis of an autocorrelation based frequency tracker for LFM and QFM signals

“…Simulations in [7] discuss the error performance of the recursive frequency estimation algorithm implemented using a forgetting factor λ , when the frequency 0 f of the signal is a constant. It has been shown that an estimator using a lag value m equal to a third of the signal length asymptotically achieves the CRB.…”

“…A simple and efficient recursive technique to implement the inverse tan operation in the PP method has been reported recently in [5] using the CORDIC (COordinate Rotation DIgital Computer) architecture [6]. The extension of the PP method to frequency estimation and tracking of signals where the instantaneous frequency (IF) is time varying is reported by us in [7]. The tracking is performed with the use of a simple, efficient chirp rate estimator implemented via state-space modeling [8].…”

“…using Field Programmable Gate Array (FPGA) [9]. In here, we provide a detailed performance analysis of the autocorrelation based frequency tracker proposed in [5,7] for using with linear frequency modulated (LFM) and quadratic frequency modulated (QFM) signals.…”