A Hybrid Algorithm for Estimating the Parameters of Polynomial-Phase Signals
Pengfei Tang,Qianqiang Lin,Bin Yuan,Zengping Chen
DOI: https://doi.org/10.2991/iccsee.2013.456
2013-01-01
Abstract:This paper details a hybrid algorithm for estimating the parameters of polynomial-phase signals (PPSs) embedded in additive white Gaussian noise. This algorithm combines the high-order ambiguity function (HAF) and the generalized cubic phase function (GCPF), and is a modification of the hybrid CPF-HAF method. In the proposed algorithm, the HAF is first applied on the original PPS to produce a cubic phase signal, whose parameters are then estimated by the GCPF, instead of the CPF. Numerical examples reveals that the signal-to-noise-ratio (SNR) threshold and mean-squared error (MSE) of the proposed approach are significantly lower than the HAF, and the SNR threshold of the proposed approach is increased by about 4dB with respect to the CPFHAF, while the MSE is reduced by about 2dB above the threshold. In addition, the proposed approach is extended to the product version in the presence of multicomponent PPSs, where the product version of the CPF-HAF suffers from an identifiability problem. Keywordsparameter estimation; polynomial-phase signal; multicomponent signals; high-order ambiguity function; generalized cubic phase function I. INTRODUCTIO The polynomial-phase signal (PPS) model is widely used for modelling signals in radar, sonar, biomedicine, communications, etc [1, 2, 3]. There are many techniques for estimating the parameters of PPSs. Of these techniques, the best known is the HAF [4]. It is a phase differentiation (PD) technique which decreases the order of the signal phase by one in each stage of the procedure. The resulting signal in the final stage is a complex sinusoid whose frequency is directly proportional to the highest-order phase parameter. With FFT algorithm and one-dimensional search, the highest-order phase parameter can be estimated. The obtained estimate is used to dechirp the original signal and the procedure is repeated until all the parameters are estimated. For an arbitrary Pth-order PPS, the HAF recursively applies PD calculation P-1 times. The procedure requires the multiplication of 1 2 signal terms, which causes reduced accuracy and numerous cross-terms when multicomponent signals are present [5]. The product highorder ambiguity function (PHAF) [3], which is the extension of HAF, reduces the noise influence and cross-terms by using the product of several HAFs calculated with various lags and proper scaling. Recently, a simple modification of HAF has been proposed [6, 7]. When dealing with monocomponent PPS, this approach indeed outperforms the HAF in terms of the accuracy and signal-to-noise-ratio (SNR) threshold. However, the approach suffers from an identifiability problem when multicomponent PPSs are considered. The generalized cubic phase function (GCPF) has been proposed for estimating parameters of a cubic phase signal [8, 9]. Its product version, product GCPF (PGCPF), is capable of dealing with multicomponent cubic phase signals. In this paper, an approach that combines the GCPF and the HAF has been proposed. In Section II, a short overview of the signal model and the HAF, PHAF, CPF and GCPF techniques is given. In Section III, the proposed GCPF-HAF approach is detailed. Extension to multicomponent PPSs is described in Section IV. In Section V, numerical simulations are carried out to evaluate the performance of the proposed approach. Finally, conclusions are drawn in Section VI. II. BRIEF REVIEW OF HAF, PHAF, CPF AND GCPF The monocomponent PPS embedded in noise can be modelled as