Target Detection And Tracking Techniques For Multi-Frequency HFSWR

The radar system OSMAR was developed by the Radio Propagation Lab of Wuhan University in the golden era of the development of high-frequency surface wave radar (HFSWR) after the Cold War. In this period, the small radar systems for ocean enviroment monitoring was the mainstream of civilian HFSWR which was represent by the CODAR's Seasonde. So OSMAR was given the mission of the marine environmental monitoring for China in its birth. However, radars are for target detection inherently, and also OSMAR. Target detection and tracking technique is one of the most important techniques of OSMAR. It does not belong to OSMAR's yesterday, it is OSMAR's future. So today, no matter in what form of OSMAR, target detection and tracking is a core technique of our study."Frequency-Agile Multifunctional HFSWR" is the latest form of OSMAR. Its appearance is not accidental. After the wave of sdudy in the radar's civilian use, the focus of HFSWR's research soon returns to the field of target detection. "Multifunctional" is the product of the times. HFSWR has its natural advantages and disadvantages in target detection. Its detection performance is instable because of sea clutter and fluctuation of target's radar cross surface (RCS). It is long ago that peaple tried to use multi-frequency to compensate for this defect, and failed to succeed more. In fact, the " Frequency-Agile Multifunctional HFSWR " is just a tool for target detection technology.This paper launch the research of HFSWR's target detection and tracking techniques on the platform of that radar system in the base of the reseach supported by the national 863 Project-" Frequency-Agile Multifunctional HFSWR ". Its contents range all the radar signal processing and data processing from the radar waveform design to the multi-target tracking, and it shows a whole picture of HFSWR's target detection and tracking technique. A careful in-depth study is made on some of the specific techniques of OSMAR's target detection and tracking, and a unique understanding of those techniques are given. Some issues left over by history are answered through distinction of concept and basic theory analysis. In addition, based on comprehensive analysis of the characteristics of the radar signal, the author proposes a set of engineering programs for target detection and tracking, and gives the results of its application, and makes a practical evaluation.The main contents are as follows:1. Through analysis of frequency modulated interupt continuous wave (FMICW), its demodulation algorithm are summarized, and the advantages of the range demodulation by direct DFT are pointed out. A demodulation algorithm based on the time-frequency analysis methods is proposed and explained from physical view. The algorithm can effectively improve signal to noise ratio and range resolution. On this basis, the author gives the overall program of the target detection and tracking algorithm, which includes channel calibration, target detection, target parameter estimation, multi-frequency data fusion, target tracking, and the demodulation algorithm memtioned before. There are 6 parts in totle. Each part of the signal processing algorithm is interrelated with others but distinct.2. The principle of active and passive channel calibration was summarized and 2 off-line active calibration solutions are proposed. Through analysis of the characteristics of actual radar signal, the author proposes a channel calibration method based on aided information from the shipborne automatic indentification system (AIS), too.3. The sea echo's characteristics are briefly introduced, and typical spectrum of target and noise are shown. For different type of radio frequency interferences, the formation mechanisms and suppression algorithms are researched carefully. The author analyzes the formation mechanism of the Doppler broadening phenomenon which often happens in the higher bands of the multi-frequency HFSWR, and proposes a signal enhancement method of velocity compensation for moving targets. On this base, this paper presents a practical constant false alarm rate (CFAR) detector.4. The estimation of direction of arrival (DOA), distance and radial speed is researched. The author studies in depth the basic theory of super-resolution algorithm, and proposes a unique view of the nature of the "super-resolution". The algorithms of MUSIC and AIC are proved applicable and put into application. From the perspective of the waveform demodulation, the author gives a suitable distance estimation algorithm for the radar which is simple and effective. On this basis, the author proposes a target parameter estimator for the multi-frequency HFSWR and evaluates it by its result.5. The basic principles of information fusion is introduced. According the characteristics of multi-frequency HFSWR, a set of plot fusion methods is proposed, including fusion of radar data from different frequencies and fusion of radar and AIS's data. The integration of AIS information is mainly for the calibration. And the multi-frequency fusion provides a single collection of target plots for tracking.6. Target tracking techniques are introduced, including multi-state model theory, the joint probabilistic data association theory and the extended Kalman filtering theory. A practical tracking program is given. The target detection and tracking algorithm is applied to the actual radar data, and its evaluation is given from the result at the end of this paper.