A Kalman-based Doppler tracking algorithm for underwater acoustic spread spectrum communications. (1st January 2022)

• Record Type: Journal Article
• Authors: Sun, Dajun; Hong, Xiaoping; Cui, Hongyu
• Is Part Of: Applied acoustics. Volume 185(2022)
• Publication Date: 2022-01-01
• Subjects: Underwater acoustic communications -- Spread spectrum -- Passband Doppler tracking -- Kalman filter; Acoustical engineering -- Periodicals; Periodicals; 620.2
• DOI: 10.1016/j.apacoust.2021.108374
• Languages: English
• ISSNs: 0003-682X

Abstract

Abstract: This paper investigates a method that dynamically tracks and compensates the Doppler spread of the spread spectrum signal in underwater acoustic communications. Due to the ultra-wideband property of the underwater acoustic signal, the Doppler spread is manifested as frequency shifting and time scaling, i.e., dilation or compression. The accurate estimation of the waveform dilation/compression in the time domain corresponds to the Doppler spread. A finer Doppler spread based on the passband signal can be discriminated due to the higher sampling rate than the baseband signal. In this paper, a novel Kalman-based Doppler tracking and compensation algorithm that operates on passband for underwater acoustic spread spectrum signal is addressed in a symbol-by-symbol fashion. The received signal is firstly correlated with the local passband reference signal to obtain the despreaded passband signal, and then a fractional time delay estimation on the passband signal by cosine waveform approximation is performed to achieve a higher accuracy of time delay estimation. The time difference of the two adjacent symbols is used as the input of the Kalman filter based on the delay-Doppler equation to estimate the instant Doppler factor sequentially. This algorithm adaptively changes the local reference signal based on the filtered Doppler factor for compensating magnitude distortion induced by the velocity variation. Both numerical simulation and experimental data analysis are presented to demonstrate the usability of the proposed algorithm.