A characteristic extraction algorithm of the ballistic missile hyperspectral data based on the quantization coding method is studied. The ballistic missile hyperspectral data of imaging spectrometer based on near-space platform is generated by numerical method. The characteristic of the ballistic missile hyperspectral data is extracted and matched using the characteristic extraction algorithm based on the quantization coding method. The simulation result show that the characteristic extraction algorithm is easy to implement and low complexity. The characteristic, which extracted by this algorithm, is enough to represent the speciality of the ballistic missile hyperspectral data. The matching algorithm has high accuracy.

This paper discusses a Web-based remote radar fault diagnosis and information system . A delicated radar network is set up with knowledge base and Web . The various key points about the structure , the function module and the key technology of the remote diagnosis system are analyzed and developed . In this way the real-time monition and automatical fault and diagnosis are realized , the reliability of the radar and the emergency repairing ability is improved .

A characteristic extraction algorithm of the ballistic missile hyperspectral data based on the horizontal traversing time is studied. The ballistic missile hyperspectral data of imaging spectrometer based on near-space platform is generated by numerical method. The characteristic of the ballistic missile hyperspectral data is extracted using the characteristic extraction algorithm based on the horizontal traversing time. The simulation result show that the characteristic extraction algorithm is easy to implement and low complexity. The characteristic, which extracted by this algorithm, is enough to represent the speciality of the ballistic missile hyperspectral data. The characteristic is not sensitive to disturbed signal and improve the timeliness of the following object matching and recognition.

A new method of detection of coding signals under the background of sea clutter is presented. The process of signal detection consists of three stages: modeling sea clutter signals based on chaos, one-step ahead prediction of chaotic signals and detection decision making. In this method, models of chaotic signals were created in the form of multi-layer perceptron neural networks, coded signals take on 13-element Barker code. The experiment results show detection of coding signal by using this method has higher detection probability and lower false alarm probability and good performance of the whole detection although with a low SNR. This method turned out to be very robust to different chaotic signals.