An algorithm for parameter estimation of multicomponent chirp signals

This paper presents an algorithm for estimating the parameters of multicomponent chirp signals. The estimator is based on the cubic phase function (CPF), which is efficient to estimate the parameters of monocomponent polynomial phase signals (PPS) with order is less than or equal to 3. When the CPF is dealing with multicomponent chirp signals, the spurious peaks arise and thus the identifiability problem occurs. A new approach based on the transformation called product cubic phase function (PCPF) is proposed to remove this problem. This estimator offers a number of advantages with respect to CPF including improved noise rejection, suppression of cross terms, and elimination of spurious peaks. The algorithm is verified by simulation results.