Improved De-Multipath Neural Network Models With Self-Paced Feature-to-Feature Learning for DOA Estimation in Multipath Environment

Xiang, Houhong; Chen, Baixiao; Yang, Ting; Liu, Dong

doi:10.1109/tvt.2020.2977894

Cited by 49 publications

(24 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…하지만, 기존의 프로 세스는 개별 소자의 빔패턴 진폭 변화로 왜곡을 판단하 므로 정확한 위상 정보가 요구되는 레이다 배열 안테나 시스템 평가에 한계가 있다 [5], [6] . 이러한 위상 왜곡 분석의 한계를 극복하기 위해 목표물 탐지 오차를 최소화하기 위한 다양한 알고리즘적 접근법이 제시되어 있으나, 연산 복잡도를 증가시키고, 특정 위상 왜곡 임계치를 넘어서면 성능 개선이 미비하므로 효율성이 낮다는 단점이 있다 [7] ～ [9] . 또한 또한, 탑재 구조체 유무에 따른 능동 소자 패턴 왜곡…”

Section: ⅰ 서 론unclassified

Performance Evaluation of Radar Array Antennas Based on Complex Received Currents Using Full-Wave Simulations with Surrounding Platforms

Yu¹,

Lee²,

Park³

et al. 2021

J. Korean Inst. Electromagn. Eng. Sci.

View full text Add to dashboard Cite

In this study, we propose a systematic performance evaluation process based on complex received currents for searching and tracking radars. This process incorporates an approach to simplified geometries of array antennas and cannon platforms that accurately represent the electromagnetic properties of the original geometries. The simplified geometries are used in full-wave electromagnetic simulations to compute the current magnitudes and phases induced in array antenna ports, and the values are saved as complex steering vectors. In addition, we propose a novel metric for evaluating the received performance distortions using the saved complex steering vectors, which were cross-checked with a conventional pattern distortion metric. To verify the feasibility of the proposed process, the optimal placement of array antennas for a commercial searching and tracking radar system was determined. The results demonstrate that average magnitudes and phases distortions of less than 1 mA and 3°, respectively, can be obtained when the array antennas are placed at 0.6and 1.1-m heights above the platform.

show abstract

Section: ⅰ 서 론unclassified

Performance Evaluation of Radar Array Antennas Based on Complex Received Currents Using Full-Wave Simulations with Surrounding Platforms

Yu¹,

Lee²,

Park³

et al. 2021

J. Korean Inst. Electromagn. Eng. Sci.

View full text Add to dashboard Cite

show abstract

“…e large range difference ΔR between the DP signal and the multipath signal results in large relative time delay Δτ, that is, Δτ > 1/2f s , where f s is the sampling frequency. erefore, the complex envelopes of the DP signal and the multipath signal will be misalignment, leading to the failure of the joint steering vector class method [27] for signal detection.…”

Section: E Modeling and Analysis Of The Synthetic Multipathmentioning

confidence: 99%

Incoherent Integration Detection Method of Airborne Phased Array Radar in a Multipath Environment

Zhao

Chen

2021

International Journal of Antennas and Propagation

View full text Add to dashboard Cite

Airborne phased array radar (PAR) suffers from multipath problems when flying over a calm sea surface. The existence of a multipath phenomenon will cause the electromagnetic echo of the same target to be reflected back to the airborne PAR from two paths, namely, direct path (DP) and multipath. Compared with the ground-based radar, the target echo received by airborne PAR in the multipath environment has two important characteristics: one is that the DP signal and the multipath signal exist in different range bins, and the other is that the radar cross section (RCS) in the DP direction may be smaller than that in the multipath direction. Considering these two characteristics, this paper first proposes a target pairing algorithm for matching the DP range and multipath range of the same target in signal detection, and then, combined with the cell-averaging constant false alarm rate (CA-CFAR) detection model, an incoherent integration detection method for airborne PAR in the multipath environment is proposed. In the target pairing process, the geometric structure relationship of the airborne PAR model can be fully utilized. After a successful target pairing process, the energy of the multipath signal will be incoherently accumulated into the corresponding DP range bin, so as to improve the probability of DP range bin data passing the detection threshold. In essence, the proposed method makes full use of multipath energy to improve the detection capability of airborne PAR in the multipath environment. Finally, the detection probability of the proposed method is given, and the detection performance is analyzed.

show abstract

“…However, the existing DL-based estimation methods are all based on supervised learning [24], [25]. During training the networks, these methods require us to attach the correct labels for training and validation datasets [26]. In reality, the actual labels (such as angle and power spectrum) are hard to collect.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised Learning Strategy for Direction-of-Arrival Estimation Network

Yuan¹,

Wu²,

Wu³

et al. 2021

IEEE Signal Process. Lett.

View full text Add to dashboard Cite

In this letter, we proposed a novel unsupervised learning strategy for direction-of-arrival (DOA) estimation network. Inspired by the sparse power spectrum and 1-norm optimization, we develop a novel loss function to cooperate with the estimation network. Unlike the prior DL-based methods, the proposed method does not need any manual annotations for training and validation datasets. Compared with state-of-art methods, the proposed method can automatically increase the degree of freedom of the array without further pre-processing on the covariance matrix of array observation data. Moreover, the proposed method can obtain clear spectrum and precise DOAs under harsh estimation environments.

show abstract

Improved De-Multipath Neural Network Models With Self-Paced Feature-to-Feature Learning for DOA Estimation in Multipath Environment

Cited by 49 publications

References 14 publications

Performance Evaluation of Radar Array Antennas Based on Complex Received Currents Using Full-Wave Simulations with Surrounding Platforms

Performance Evaluation of Radar Array Antennas Based on Complex Received Currents Using Full-Wave Simulations with Surrounding Platforms

Incoherent Integration Detection Method of Airborne Phased Array Radar in a Multipath Environment

Unsupervised Learning Strategy for Direction-of-Arrival Estimation Network

Contact Info

Product

Resources

About