期刊
APPLIED ACOUSTICS
卷 171, 期 -, 页码 -出版社
ELSEVIER SCI LTD
DOI: 10.1016/j.apacoust.2020.107557
关键词
Orthogonal frequency division multiplexing (OFDM); Doppler estimation; Bit error rate (BER); Gradient optimization
类别
资金
- National Key Research and Development Program of China [2018YFE0110000]
- National Natural Science Foundation of China [11274259, 11574258]
This paper proposes an iterative Doppler estimation algorithm based on bit-error-rate (BER) gradient descent strategy to reduce search complexity, with smoothed gradient and variable Doppler increment to improve convergence rate and avoid local optimums. The effectiveness and superiority of the proposed BER gradient descent Doppler estimation algorithm is verified through numerical simulations and sea trials.
With its robustness to multipath spread the Orthogonal frequency division multiplexing (OFDM) has drawn rapidly increasing interests as a high data rate underwater acoustic (UWA) communication solution, which is unfortunately subject to significant performance degradation caused by Doppler. Generally Doppler can be estimated by searching the best match between the received packet and the standard copy in time domain or frequency domain at signal-level. However, under dynamic or diffuse multipath in adverse UWA channel there may exist mismatch between the signal-level Doppler estimation and bitlevel demodulation. To address this problem, a bit-error-rate (BER) based Doppler estimation strategy has previously been derived by searching the optimum BER among Doppler grids at the expense of high complexity caused by exhausting search. In this paper by adopting BER gradient descent strategy an iterative Doppler estimation algorithm is proposed to substantially reduce the BER search complexity. Moreover, to improve convergence rate as well as to avoid being trapped at local optimums, smoothed gradient and variable Doppler increment are adopted in the iterations. Finally the numerical simulation as well as sea trial result is provided to verify the effectiveness and superiority of the proposed BER gradient descent Doppler estimation algorithm. (C) 2020 Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据