Study On Ionosphere Interference Suppression Techique Based On Convex Optimization

The detection performance of HFSWR, an important part of the coastlinewarning system, influences the security of our country and people directly. TheIonosphere Clutter that is the biggest obstacle of influencing the ability of HFSWRdetection, has been receiving extensive attention of researchers in the world. Its time-varying and instability characteristics restrict its further research and its suppressionmethod has been an international problem for a long time. Recently, the technique ofconvex optimization is continuously applied in the field of signal processing, becauseit can provide the best optimizing solution of certain error range which can be usedto get a stable optimal solution for the problem of the Ionosphere Clutter that has acertain non-stationary characteristics. For this reason, the technique of convexoptimization may provide a new direction of researching on the problem ofionosphere interference suppression.This dissertation proposes a high performance and strong practicabilitytechnique of Ionosphere Clutter suppression that can protect the targets from losing,enhance the output ratio of signal to noise and strengthen the detection performanceof HFSWR, at the meantime ensuring the anti-disturbance performance, onIonosphere interference suppression combined with the data of actual measurementfrom radar system by analyzing the characteristics of Ionosphere Clutter. Theachievements of this dissertation mainly comprise the following points:1. Consider a self-revised robustness adaptive beam-forming algorithm for the non-stationary Ionosphere Clutter. This algorithm operates the sampling matrix whileconsidering the rectification of steering vector. This operation eliminates theexpected target information and gets a sampling matrix that only contains theinformation of interference and noise. This kind of sampling matrix is a noise andinterference covariance matrix, so it has an excellent performance of anti-interference compared with traditional algorithm.2. Have a deep research on the characteristics of Ionosphere Clutter and discussionon its characteristics of time domain, range domain, Doppler domain and spacedomain by using measured data. Achieve the feature of relatively strongIonosphere Clutter range correlation, get some knowledge about the distributionof Ionosphere Clutter space domain and have a comparison among the spacedomain features of targets, sea cluster and Ionosphere Clutter.3. Propose an effective technique of convex optimized Ionosphere Cluttersuppression starting with the selection of training samples. The algorithmcomputes the space spectrum used to the measured data by sparse decomposing and makes full use of the differences between the targets and cluster spacespectrum to separate and reconstruct the data of targets and cluster. Have theoperation of range correlation suppression and space domain filtering on the dataof cluster at same time, and then get the handling result by fusing with the targetdata. The algorithm has a role of suppression on many kinds of Ionosphere Clutterand it can detect the targets submerged by clusters to improve the background oftarget detection. Meanwhile, the algorithm has high parallelism meaning that itcan be computed by using massively parallel computing devices which providesthe theoretical guarantee of the real-time of engineering implementation. Theprocessing result of measured data indicates that the algorithm has a highperformance and it indeed suppress the Ionosphere Clutter and enhances theprobability of detecting the target.