本研究主要為開發一套全新自動化辨識與濾除射頻干擾(Radio Frequency Interference，RFI)及電離層干擾(Ionospheric Interference)演算法並應用至高頻地波雷達，內容包含海面回波特性、操作波形(Operational Waveform)、來向角(Direction of Arrival，DOA)演算法等介紹，利用互頻譜分析(Cross Spectral Analysis)統計發現約8成觀測時間內頻譜被干擾影響，經干擾濾除後與浮標提供徑向洋流都卜勒速度(Radial Current Doppler Velocity)比對，相關係數(Correlation Coefficient)最佳提升至0.82，且方均根誤差(Root Mean Square Error)最佳下降至0.42m/s。本研究首次改良並應用頻率域自適應雜波抑制法(Frequency domain Adaptive Clutter Suppression，FACS)於高頻地波雷達射頻干擾，經統計後發現當參考距離閘數目為2時，海波訊雜比(Signal to Noise Ratio，SNR) 最佳達到約4.06dB改善，且射頻干擾濾除後仍保有目標物來向角特性，並使一定數量干擾飽和(Saturation)頻譜重新獲得海波。此外本研究也是首次將頻率域自適應雜波抑制法改良並應用至電離層探測儀電離圖中，並與射頻干擾抑制法(Radio Frequency Interference Mitigation，RFIM)比較與分析，得出RFIM標準差倍數門檻為2-2.5倍及FACS參考距離閘數目為5-15個有較佳結果。因為調頻截斷連續波(Frequency Modulated Interrupted Continuous Wave，FMICW)波形設計，高頻地波雷達中電離層干擾易有距離混疊效應(Aliasing)，透過電離層探測儀(Ionosonde)電離圖(Ionogram)進行距離比對，經距離混淆修正後，相關係數最佳提升至0.99，方均根誤差最佳下降至2.03km。本研究改良並應用頻率域正交投影法(Orthogonal Projection Algorithm，OPA)於高頻地波雷達電離層干擾，並新增方法使用三維雷達資料，其中經統計得知電離層干擾相干值(Coherence)較海洋回波高，可用於辨識不同回波之門檻。最後因為特高頻(Very High Frequency，VHF)測海雷達頻譜不易受外部干擾影響，利用其資料進行徑向洋流都卜勒速度、示性波高(Significant Wave Height)等測海參數及海面目標物來向角反演，並利用漂流浮標(Drifting Buoy)、自動識別系統(Automatic Identification System，AIS)及廣播式自動回報監視(Automatic Dependent Surveillance-Broadcast，ADS-B)等資料進行比對，驗證測海雷達參數反演與海面目標物來向角計算演算法正確性，經資料品質控制(Quality Control，QC)與相位修正後，其結果皆有良好正相關性。;In this study, we develop a new automatic algorithm for high frequency ground wave radar (HFGWR) to identify and mitigate the influence of radio frequency interference (RFI) and ionospheric interference. We also introduce characteristic of sea echo, operational waveform, and estimation of direction of arrival (DOA). Use cross spectral analysis to statistic chatacteristics of interference. It shows there are over 80% spectrum contain interference during experiment. After mitigation of interference, compare radial current Doppler velocity with drift. The best results show that correlation coefficient increase to 0.82 and root mean square error decrease to 0.42m/s. In this study, we first improve and apply frequency domain adaptive clutter suppression (FACS) for HFGWR to mitigate RFI. It shows the best improvement of SNR is 4.06dB when the number of neighboring range cell is 2. DOA is consistent after mitiagation of RFI. It also shows there are some saturated spectrum recovering the information of sea echo. We also first improve and apply frequency domain adaptive clutter suppression (FACS) for iongram to mitigate RFI and compare the results with radio frequency interference mitigation (RFIM). The results show well when threshold of standard deviation for RFIM equals to 2-2.5 and number of neighboring range cell of FACS equals to 5-15. Limited by characteristic of Frequency Modulated Interrupted Continuous Wave (FMICW), range aliasing of ionospheric interference sometimes occur for HFGWR. Use ionogram provided by ionsonde to compare range. After revising of range aliasing, the best results show that correlation coefficient increase to 0.99 and root mean square error decrease to 2.03km. In this study, we first improve and apply orthogonal projection algorithm (OPA) for HFGWR to mitigate ionospheric interference. We add a method to use three dimensional radar data. Coherence of ionospheric echo is higher than that of sea echo. This characteristic can be used to set thrsohold to identify different echoes. Finally, very high frequency (VHF) sea radar has less effect by external interference. Estimate radial current Doppler velocity, significant wave height and DOA from target to compare with drift, automatic identification system (AIS) and automatic dependent surveillance-broadcast (ADS-B). Validate the algorithm for retrieval of sea parameter and DOA of target. The results show high positive correlation after quality control and revising phase offset.
