Research On Methodlogy Of Sea Clutter Suppression In Shipborne Surface Wave Radar

Ship-borne surface wave radar is known for its excellent mobility, flexibility and good survivability. As marine resources become more and more improtant, Shipborne surface wave radar, which can be used for long distance detection, tracking, early warning, exploiting marine resources, marine protection and maritime enforcement, has drawn much more attention. Due to the movement of the platform, comparing with land based ground wave radar, the first-order sea clutter echo spectrum will be broadened, which will submerge the targets with low velosity such as ships and make it difficult for target detecting. Research on methodlogy of sea clutter suppression in shipborne surface wave radar has a great significance.This paper is mainly about the research on methodlogy of sea clutter suppression in shipborne surface wave radar. Several conventional sea clutter suppression algorithms are simulated and analyzed, and two improved algorithms are proposed. Firstly, the background of shiborne surface radar detection is analyzed, the formation mechanism of the sea clutter and the mathematical model of the power of the sea clutter echo is given. When the platform is moving, the sea clutter return from different azimuths will cause different doppler shifts. And radar’s beam width cannot be ignored, thus the sea clutter echo’s spectrum is broadened. In a word, sea clutter echo can be seen as an independent interference which has a certain direction within a radar cell. This is also the foundamention of the clutter suppression algorithms in space domain.Sea clutter suppression algorithms in space domain are widely used now. There are mainly two methods, orthogonal weighting and oblique projecting algorithms. Both of these algorithms suppress clutter by separating signal space and clutter space through projection transformation. For orthogonal weighting algorithm, suppression of interference coming from within the main beam pointing to target would result in the main beam splitting into two beams, which is unfavourable for target detection and estimation. Oblique projection can relieve this problem by projecting the received signal onto the subspace which is paralell to the signal space. In practical applications, the direction of the target is unknown and the scanning steerint vector is used to replace the target azimuth, which will affect the performance of the oblique projection algorithm.Considering the space-time coupling properties of the first-order sea clutter, it will not get a good performance for suppressing the sea clutter in only one domain. Space-time adaptive processing(STAP) which has been used in the airborne radar was applied to the shipborne HFSWR. Considering the characteristics of sea clutter echoes and the data model in the shipborne surface wave radar, two reduceddimension STAP methods, joint domain localized algorithm and space-time multiple beams algorithm, are introduced here. These two methods both transform the data to the angle-Doppler domain, then utilize the secondary data from range cells surrounding the range cell of interest to estimate the sea clutter covariance matrix. When the covariance matrix of the sea clutter is got, the space-time weighting vector can be calculated. The sea clutter is suppressed under the maximum signal to noise ratio criterion.For joint domain localized algorithm, its performance depends heavily on the accuracy of the covariance matrix of the interference. A modified calculation of the covariance matrix based on the distribution and characteristics of sea clutter is proposed to fit for sea clutter better and make maximum use of the few secondary training data. To get a better estimation of the direction of the target, a space-time cascaded processing algorithm combining advantages of spatial projection algorithm and space-time adaptive processing algorithm is proposed. This method can give a better estimation of the direction of the target while maintaining the good performance in the Doppler domain. These two methods have a better performance which is tested from measured data.