On Detecting of Linear Frequency-Modulated Waveforms in Frequency-Dispersive Channels

Abstract

Linear-frequency-modulated (LFM) waveform is usually applied in active sonar and radar systems as it has the advantage of wide bandwidth while keeping the envelope constant and pulse duration short. However, the transmitted signal of radar and sonar systems is usually distorted under frequency-dispersive channels due to the time-varying characteristics of medium, which can be modeled by the classical Fast Fading Distortion (FFD) model. In this paper, based on the FFD model, we propose a new detection algorithm based on orthogonal decomposition in order to improve the robustness of detection in the situation that the coherence time of channels is unknown. More precisely, in the proposed method, we select the frequency-shift replicas of transmitted waveform as basis and determine the number of the replicas with which we correlate the echo by measuring the estimation error of echo’s power. Furthermore, we extend the method to the case of range-distributed target and propose a low-complexity method. Numerical examples show the proposed method can give better performance when the estimation of the coherence time of the channels is not precise.