.png){kind=icon}
.png){kind=icon}
Return
Bridging the Domain Gap Between Target Detection and Parameter Estimation: A Time-Frequency Parameter Estimation Scheme for HFSWR
.png){kind=icon}
PRE
AI
DOI：10.1109/TRS.2025.3638995.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Time-frequency analysis (TFA) serves as an effective technique in improving target-monitoring performance of high-frequency surface-wave radar (HFSWR). However, the potential of this technique has not been fully explored in existing schemes because the extracted time-frequency (TF) signatures are solely used for identifying targets. In addition to missed detection, target nonstationarity may also degrade parameter estimation accuracy. To address this challenge, we propose a TF-based parameter estimation scheme. In particular, we develop range and direction-of-arrival (DOA) methods that directly leverage TF signatures obtained during the detection stage to extract target parameters. These parameters are then processed for subsequent plot association and target localization stages. Statistical results on measured data show that the proposed range and DOA estimation methods outperform their conventional counterparts in accuracy. Next, these methods are integrated into the proposed parameter estimation scheme, which is then compared against existing schemes. Experimental results demonstrate that our scheme achieves better plot association performance compared with other conventional schemes. In addition, using automatic identification system (AIS) records as ground truth, our scheme achieves enhanced localization accuracy. In particular, it reduces the proportion of anomalous coordinate trajectories by 2.05%similar to 4.11%. Moreover, by seamlessly connecting target detection and parameter estimation stages within the TF domain, this scheme streamlines the overall target-monitoring pipeline.
Keywords:
High-frequency surface-wave radar (HFSWR)
sparse iterative covariance-based estimation (SPICE)
target parameter estimation
time-frequency analysis (TFA)
Journal
I
IF:
0
Papers:
58
Citations:
0
Organization
View more organizations