Study On Low-angle Altitude Measurement In VHF Radar

As a kind of meter-wave radar, it has good anti-stealth effect, anti-arm capacity, lessattenuation compared with microwave radar, and large detection range. In recent years,the research in VHF radar has been attached great importance. Limited by aperture,meter-wave radar has wide beam, which usually causes bad angular resolution.Furthermore, high sidelobe excludes good anti-interference capacity. Because of thebeam wideness, the beam often reflects to the ground when radar detects the long rangetargets, the received signal in the mainlobe includes the direct path echo from the targetdirectly and the reflection multipath echo which are reflected from the ground (sea)surface. The reflection multipath echo cause the measurement accuracy of elevationdeteriorates sharply. The key technology to be solved is improving the low-elevationangle accuracy.On the basis of the achievements obtained by now, combined with my researchproject, this dissertation starts from low angle estimation in meter-wave radar. The maincontent of this dissertation is summarized as follows.1. Considering the demand and technical characteristic of radar signal processingsystem, a new high-powered radar signal processing system architecture based onFPGA and DSP is proposed. The architecture is based on cPCI standard platform.High-speed serial channel improves data Transmission rate. The crosspoint switch isable to make circuit boards connection dynamically. These greatly improve theperformance of the radar signal processor. This system is simple in structure and hasgood universality and extendibility which is in agreement with the forthcomingdirection of radar signal processors and has high engineering value. Additionally, theangle measurement method with digital monopulse for2-Dimensional digital arrayradar is studied in the paper. The variations of equivalent aperture of antenna array withdifferent frequencies and different beam positions are taken into account in the processof difference patterns synthesis. Multi-order polynomial fitting method is adopted tocalculate the target’s angle according to the angle error signal whose characteristic isnonlinear. This technique decreases the number of monopulse error cures and simplifiesthe huge data storage and the calculation of lookup table, facilitated engineeringimplementation. It is important to eliminate the quantized error of error curve tables anddecrease the measuring error of linear approximation. 2. A generalized constrained MUSIC is studied for low-angle altitude estimation.Constrained MUSIC improves the estimation precision of unknown source directions byincorporating the prior known source directions and has no influence caused bycoherence of sources. A generalized constrained MUSIC is proposed, which is modifiedin terms of the low-angle environment of VHF radar, improves the estimation precisionof direct signal with multi-path signal as a constraint. The proposed method deal withcoherent sources directly and doesn’t considering the affect caused by reflectioncoefficient. According to the characteristic of height-finding, some improvement ismade on the proposed algorithm, change2-D search to1-D search, which reducescomputation complexity and gets good performances. The proposed method obtainsgood performance while the number of snapshot is more than the number of elementcompared with the spatial smoothing MUSIC.3. A synthetic steering vector superresolution algorithm based on a highlydeterministic multipath signal model is proposed. To model the signal more accurately,signal polarization, mirror reflection, dispersion, diffuse reflection, the terrainparameters of the reflection region and grazing angle must be taken into account, inaddition to antenna height, distance between target and array, the curvature of the signalpath due to refraction in the troposphere and the curvature of the earth itself. Thecalculation methods of the reflection coefficient and lengths of the direct path andindirect path are given. A synthetic steering vector superresolution algorithm based onterrain matching is proposed. The accuracy DOA is obtained by superresolutionalgorithm that makes use of the synthetic steering vector which is the sum of directsteering vector and indirect steering vector. In additional, if the roughness dissatisfiesthe Reileigh rule, the received signal includes the direct component and diffusereflection component. The reflection surface is rougher; the power of diffuse reflectioncomponent is stronger. The diffuse reflection component in signal module can be takenfor coherent distributed source. The DOA could be obtained by multi-dimensionalternating projection and synthetic steering vector superresolution algorithm. But inpractice, it is hard to ascertain the multipath distributed source model that should bevalidated by lots of experiments.4. Making use of the sparsity of targets, some novel direction-of-arrival models forlow-angel estimation in VHF radar based on compressive sensing are proposed.Covariance matrix compressive sensing, interpolated array compressive sensing andbeam space compressive sensing carry out compressive sampling on covariance matrix,interpolated array and beam space, respectively, and three new three new direction-of-arrival models based on compressive sensing are built. The accuracy DOAestimation is achieved by sparse reconstruction algorithm. The proposed methodsachieve resolution beyond Rayleigh limit, get low sidelobe in spatial spectrum andobtain good performance at low SNR and less snapshots environment. The proposedmethods are conceptually different from subspace-based methods and provide highresolution using a uniform linear array without restricting requirements on the spatialand temporal stationary and correlation properties of the sources and the noise and gethigh angular resolution.5. An interferometric-like APES algorithm is studied for DOA estimation ofdistributed coherent arrays. The number of targets and coarse direction cosine estimatesare obtained from interferometric-like APES spatial spectral and model-order selectioncriterion. The high accuracy direction cosine estimates are derived from phase centershift between subarrays. A dual-size algorithm is used to resolve the ambiguity in DOAestimation resulted from the distributed coherent arrays. Then the low variance andunambiguous DOA estimates are achieved. The proposed approach is a blind DOAestimation method and is independent of the number of sources and the coherence ofsources and has good performance while the sample is less. The algorithm enhancesangle resolution and accuracy at the cost of huge computation and can be applied inVHF radar low-angle DOA problem. Also the baseline ambiguity threshold and SNRambiguity threshold in distributed coherent arrays are demonstrated, the computationcost and CRB of the proposed method are analysed.On the basis of simulations and analysis, the real data of some VHF radar isincorporated into these methods above. Therefore, this dissertation realizes the closerelation between the theory and practice or implementation, and solve some practicalissues of the VHF radar in some typical positions.This dissertation is of vast importancefor the research and development of this VHF radar.