期刊
IEEE TRANSACTIONS ON SIGNAL PROCESSING
卷 69, 期 -, 页码 4204-4218出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TSP.2021.3095722
关键词
Direction-of-arrival estimation; Smoothing methods; Array signal processing; OFDM; Estimation; Time of arrival estimation; Signal processing algorithms; Carrier frequency offset (CFO); time of arrival (TOA); direction of arrival (DOA); large-scale array
资金
- NSFC [61960206005, 61971136, 61803211, 61871111]
- National Key R&D Program of China [2020YFB1806603]
- Jiangsu NSF [BK20191261, BK20200820, BK20200393]
- Fundamental Research Funds for the Central Universities [2242021k30043, 2242021k30053]
- Open Research Fund Program of LIESMARS [20P01]
- Research Fund of National Mobile Communications Research Laboratory
- Zhishan Youth Scholar Program of SEU
This paper introduces a novel scheme for joint TOA and DOA estimation with CFO compensation using large-scale receive antenna arrays. The proposed scheme balances pilot length and estimation performance, providing insights into resolution criteria. Numerical results show that the scheme efficiently eliminates CFO effects and outperforms traditional MDL methods in low SNR conditions.
This paper investigates estimation of direction of arrivals (DOAs) and time of arrivals (TOAs) for a multi-user orthogonal frequency-division multiplexing (OFDM) system that is equipped with large-scale arrays and interfered by carrier frequency offsets (CFOs). We propose a novel scheme for joint TOA and DOA estimation with CFO compensation by taking advantages of large-scale receive antenna arrays. The pilot is designed to balance the pilot length and the estimation performance. The proposed CFO estimation scheme is with lower complexity. After analyzing the resolution of joint DOA and TOA estimation, we find a criterion to choose an appropriate smoothing window for the best resolution and provide insights in sub-carrier selection strategy. We also prove that the large-scale array can asymptotically achieve perfect resolution of both TOA and DOA estimation. We reveal the pseudo peak (PP) elimination property of large-scale arrays and exploit it to develop a novel path number estimation algorithm accordingly. Based on our numerical results, the proposed scheme can efficiently eliminate the CFO effect, the optimized smoothing window attains much better performance, and the path number estimation algorithm performs better than the traditional minimum description length (MDL) approach, especially in the low signal-to-noise ratio (SNR) regime.
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