Small ship detection with high frequency radar using an adaptive ocean clutter pre-whitened subspace method

We present a new direction finding scheme for slow targets embedded in sea clutter that combines subspace methods (MUSIC and Root-MUSIC) and time domain sea clutter and noise suppression. The ocean surface behaves as a distributed source in contrast to ships that are point sources. By mapping data to eigenspaces, the sea clutter level decreases due to its non-deterministic behavior while point targets' levels remain unchanged while enhancing estimation performance. Since subspace methods have a higher threshold and are degraded heavily by the correlated sea clutter, clutter and noise suppression is introduced to enhance algorithm performance. Both simulated and real ship targets are used to show the lower threshold and higher estimation performance of the proposed algorithm.

Within the Straits of Florida, a strongly sheared western boundary current interacts with waves and wind to produce a complex environment that is difficult to forecast, and makes in situ data collection at the ocean surface challenging. For these reasons, ensuring accurate remote measurements of current velocity, significant wave height and wind direction from high frequency (HF) radar is an ongoing priority. In this study, we deploy a subsurface mooring in the HF radar footprint that is equipped with Nortek’s Acoustic Wave and Current Profiler (AWAC), and collect wind data from a nearby meteorological station. We evaluate the HF radar measurements through a comparison with these independent in situ datasets. With regard for their sampling differences, our results show that the remote and in situ instruments compared favorably. For the majority of the deployment, the current velocity exceeded 2 m s-1 at the mooring; this created excessive drag on the buoy leading to instrument tilt. For this reason, the mooring design and instrument performance are examined in closer detail. These preliminary results demonstrate the importance of collecting remote high resolution data from HF radar over the Straits of Florida.

High-Frequency (HF) Radar is an instrument using radio waves to measure ocean currents and waves remotely. This technology has many advantages, including has unprecedented spatial and temporal resolution, can operate in any weather condition, and is not dangerous for the environment. However, HF Radar's research is still limited in Indonesia. This research aimed to analyze the tidal and residual current in the Bali Strait in July 2020. Radial velocity from two HF Radar sites is combined to obtain the total currents. Current data from HF Radar were compared with Acoustic Doppler Current Profiler (ADCP) data to investigate its accuracy. Surface current data were analyzed using harmonic analysis to separate tidal and residual currents. Comparison between HF Radar and ADCP data are in good agreement for meridional current with a very high correlation of 0.813 and a small RMSE value of 0.22 m/s. Harmonic analysis shows that the dominant currents are tidal currents. The current direction was northward (southward) at flood (ebb), with maximum northward (southward) velocities are 2.17 m/s (2.97 m/s), respectively. The residual current has a random pattern, slightly faster northward than southward, and has similar spectral with the wind.

Both wind speeds and wind directions are important for predicting wave heights near complex coastal areas, such as small islands, because the fetch is sensitive to the wind direction. High-frequency (HF) radar can be used to estimate sea surface wind directions from first-order scattering. A simple method is proposed to correct sea surface wind vectors from reanalysis data using the wind directions estimated from HF radar. The constraints for wind speed corrections are that the corrections are small and that the corrections of horizontal divergences are small. A simple algorithm for solving the solution that minimizes the weighted sum of the constraints is developed. Another simple method is proposed to correct sea surface wind vectors. The constraints of the method are that corrections of wind vectors and horizontal divergences from the reanalysis wind vectors are small and that the projection of the corrected wind vectors to the direction orthogonal to the HF radar–estimated wind direction is small. The impact of wind correction on wave parameter prediction is large in the area in which the fetch is sensitive to wind direction. The accuracy of the wave prediction is improved by correcting the wind in that area, where correction of wind direction is more important than correction of wind speeds for the improvement. This method could be used for near-real-time wave monitoring by correcting forecast winds using HF radar data.

HF skywave `over-the-horizon' radars experience exceptionally high levels of clutter from the earth's surface as a consequence of their `look-down' propagation geometry. While Doppler processing is effective in separating aircraft echoes from the clutter, thereby enabling detection, the same does not always hold for slow moving targets. In particular, ship echoes occupy the same region of Doppler space as the sea clutter, so the likelihood of target echo obscuration is high. By exploiting the known dependence of the characteristic spectral form of the clutter Doppler spectrum on radar frequency and other parameters, HF radars can maximise the prospects for detectability against particular ranges of target parameters. Nevertheless, even assuming that a radar makes use of all available a priori knowledge to optimise its parameters, small ships and boats often escape detection because their echoes are still masked by clutter in the same Doppler subspace. This is particularly likely to occur if the vessels concerned wish to remain undetected, and hence use the properties of sea clutter to their advantage. In this paper we examine the feasibility of a novel approach to the problem of detection against sea clutter -- detection via nonlinear scattering mechanisms. The nonlinear scatterers of concern here include electronic systems as well as metal-oxide-metal junctions formed by corrosion in the marine environment. This so-called `rusty bolt' effect has previously been considered in the context of electromagnetic compatibility of communication systems. Here we explore the possibility of exploiting it to achieve a form of `clutter-free' radar.

Simulation of constituent transport using a reduced 3D constituent transport model (CTM) driven by HF Radar: Model application and error analysis

Track-Before-Detect (TBD) is a detection technique that simultaneously tracks and detects a target. The technique is specially useful for targets whose back-scattered return is significantly lower than the surrounding noise and background environments. Different algorithms proposed to perform TBD include particle filter, dynamic programming and Hough transform algorithms have been proposed. These techniques work well in a highly noise environment with constant velocity targets. However, some of these techniques are not applicable to targets that have acceleration as well as targets embedded in sea clutter. In this paper, we present a beamlet track-before-detect technique for maritime detection. Due to the sea clutter, two different adaptive sea clutter filtering methods are applied before the aforementioned TBD technique. Overall, the paper introduces and studies a beamlet TBD technique for maneuverable maritime targets in sea clutter.

Maritime Conditions represent a challenging environment for radar detection due to the so-called sea clutter. It may lead to a lower probability of a correct target detection. Thus, a detailed modelling of sea clutter with its time-dependent correlations is crucial for evaluating the robustness of modern radar systems and their signal processing. This paper utilizes a stochastic approach to generate time-correlated sea clutter, dependent of the used radar system, the environment and the measuring geometry. Therefore, different models are used to determine the RCS of the illuminated sea surface, the time-dependent correlations and finally the statistical distribution of the received signal-amplitudes. Furthermore, bistatic radar geometries should be at the scope of this paper. Afterwards the simulated signal data is compared and validated with free accessible, recorded radar data from different locations and environmental conditions. The comparison with the correlation properties and probability density functions of the radar data shows good accordance with the data generated by the model. Furthermore the results show that generating sea clutter with the developed stochastic model and a minimum of basic observeable and available quantities (sea state, wind direction) is sufficient accurate and possible. By utilizing the presented method, time-correlated radar data, which imitates the different properties of real-world sea clutter, is generated in an efficient, less time consuming and reliable laboratory environment with a minimum of known environmental parameters.

Test, measurement and verification of radar are necessary to ensure function according to specification and meet customer requirements. Extensive field tests are often required to test an entire radar system. This is not only expensive, but also requires a great deal of time. To reduce test complexity and cost, simulation techniques are used to recreate real-world conditions in a laboratory environment. Many subsystems can be tested by radar target generators simulating radar echo signals. They usually consider Swerling models, but struggle to simulate clutter, or in particular sea clutter. This paper introduces an approach to generate sea clutter with a state of the art vector signal generator. It shows common approaches modeling sea clutter and introduces a novel approach to combine the statistical models with a physical ocean surface model. This way, time and spatially correlated data that reproduces real-world sea clutter is generated, thus a cheaper, more time efficient way to conduct sea trials is presented.

Traditionally, radar signals have been analysed in either the time or the frequency domain. Joint time–frequency representations characterise signals over a time–frequency plane. They thus combine time-domain and frequency-domain analyses to yield a potentially more revealing picture of the temporal localisation of a signal's spectral components. Therefore, for air target returns with time-varying frequency content, joint time–frequency representations offer a powerful analysis tool. A concise review of time–frequency transforms is provided as background and is needed to appreciate how time–frequency processing methods can improve conventional time or frequency processing methods. The authors use time–frequency analysis techniques for the detection of a manoeuvring aircraft using high frequency (HF) radar in heavily cluttered regions. They compare the ability of different time–frequency transforms to resolve several experimental aircraft returns. The relative speeds of the different transforms are also quantitatively studied. The results clearly demonstrate that time–frequency analysis techniques can significantly improve the detection performance of the HF radar and add considerable physical insight over what can be achieved by conventional Fourier transform methods currently used by HF radars.

Improved understanding of the wave-current interaction can lead to better descriptions of ocean wave or wind and more precise designs of ocean engineering. High-frequency (HF) radar, as an important ocean remote sensing tool, can measure large-area currents and wave parameters simultaneously. Here, the first-order spectrum power (FSP) of HF radar is used to study the wave-current interaction. The frequency shift of the first-order spectrum is combined with its power to indicate the relationship between current and wave. The effects of current on the Bragg wave in different sea areas are presented by analyzing the FSP change with current vector (FSP-current distribution). The FSP-current distribution in deep water is very similar to the theoretical distribution of two-dimensional wave-current; while, the wave-current interaction over a relatively shallow shelf is stronger and more complicated. Experimental results illustrate the potential of HF radar in studying the wave-current interaction and show that the wave-current interactions in deep water mainly belong the two-dimensional oblique model (refraction case).

Abstract. Part I of this history describes the motivations for developing radars in the high frequency (HF) band to study plasma density irregularities in the F region of the auroral zone and polar cap ionospheres. French and Swedish scientists were the first to use HF frequencies to study the Doppler velocities of HF radar backscatter from F-region plasma density irregularities over northern Sweden. These observations encouraged the author of this paper to pursue similar measurements over northeastern Alaska, and this eventually led to the construction of a large HF-phased-array radar at Goose Bay, Labrador, Canada. This radar utilized frequencies from 8–20 MHz and could be electronically steered over 16 beam directions, covering a 52∘ azimuth sector. Subsequently, similar radars were constructed at Schefferville, Quebec, and Halley Station, Antarctica. Observations with these radars showed that F-region backscatter often exhibited Doppler velocities that were significantly above and below the ion-acoustic velocity. This distinguished HF Doppler measurements from prior measurements of E-region irregularities that were obtained with radars operating at very high frequency (VHF) and ultra-high frequency (UHF). Results obtained with these early HF radars are also presented. They include comparisons of Doppler velocities observed with HF radars and incoherent scatter radars, comparisons of plasma convection patterns observed simultaneously in conjugate hemispheres, and the response of these patterns to changes in the interplanetary magnetic field, transient velocity enhancements in the dayside cusp, preferred frequencies for geomagnetic pulsations, and observations of medium-scale atmospheric gravity waves with HF radars.

This paper proposes a procedure for simulation of the temporally and spatially correlated low grazing angle sea clutter based on the statistical and correlated characteristics. The magnitudes are Pareto distributed and the temporal correlation characteristics in separated range bins are different from each other. Firstly, the memoryless nonlinear transform (MNLT) is adopted to directly simulate the Pareto distributed high resolution range profiles (HRRPs) of the temporally and spatially correlated sea clutter from coherent pulses. Secondly, by utilizing the approximately linear relationship between the central Doppler frequency and the magnitude of low grazing angle sea clutter, Doppler spectrum of each range bin is modeled. Then, the alternating projection (AP) approach is used to retrieve the phases of sea clutter with desired Doppler spectrum and given magnitudes. Finally, simulation results show the Pareto distributed temporally and spatially correlated sea clutter with desired spatial correlation function and range resolved Doppler spectra.

Land-based High Frequency (HF) surface wave radars have become an effective tool for sensing currents, winds and waves, and many types of HF radars are in operation around the world. Being different from land-based HF radars, this paper proposes a concept of shore-to-air bistatic HF radar for ocean observation. The configuration of shore-to-air bistatic HF radar is introduced in the paper, and the first-order scattering model is also presented. The paper studies and analyzes the average scattered cross section per unit area at different operating frequencies and with various wind speeds via simulation. Simulations demonstrate that the average scattered cross section per unit area varies along with frequency, wind speed and scattering direction. The proposed model will help develop shore-to-air bistatic HF radars for ocean observation.

HF (high frequency) coastal radar system is used to capture the surface current behavior — in terms of velocity and direction — in the ocean near the coast. 18 HF coastal radar stations were implemented along the Gulf of Thailand in order to monitor for disasters (e.g., Tsunami) as well as relevant risks. The HF systems are also to serve other life-critical applications, such as water quality control and monitoring, chemical spill backtracking, and marine navigation. However, not all the applications can benefit from this near-real-time HF data; some applications in different domains require forecast values. The examples include search-and-rescue system and hazardous materials spill trajectory prediction. Therefore, in this paper, we propose a predictive model for future current data based on historical HF coastal radar data sets, utilizing association rule mining combined with an object dispersion concept. So, the full potential of HF radar systems can be exploited. The spatial and temporal dimensions are taken into account when designing our predictive system, which consists of two phases: ocean surface current track formulation and spatio-temporal association rule mining. The experiments are performed on a two-year HF radar dataset (2014–2015) using Google Cloud Platform. The resulting forecast current values: velocity and direction are then compared with testing datasets (using 10-fold cross validation) of the actual recorded values and evaluated based on percentage accuracy and RMSE, respectively.