Research On The Technology Of Track Sequential Connection And Target Matching Identification Based On Narrow Band Radar Network

Radar target recognition technology refers to determine the attribute, category or type of the target with analyzing the information of radar echo. The conventional air-defense warning radars in China are generally narrowband radars with low resolution. Active, narrow-band and low-resolution characters determine that this kind of radars cannot get enough information to precisely distinguish the type of the target in the process of target recognition. With the increasingly urgent demand of recognizing unknown aircraft on the border, narrowband radar as the battle equipment in our country, must be fully excavated its potential on the target recognition as much as possible, tuned combat effectiveness to maximize performance and provide powerful information support for operation decision. This paper takes two steps to achieve the target recognition of the narrow-band radars. At first, the information that narrowband radar gets is inaccurate and incomplete. It is manifested as follows: system platform ignores the problem of location inaccuracy that is caused by earth curvature; the problem of height inaccuracy of distance early-warning radar(for exampe, meterwave radar); the problem of uncontinuous track that is caused by low-altitude location of the target or the cloud by other ground-objects. How to effectively integrate target information is important to target matching recognition of narrow-band radars. In this paper, we introduce the geographic information data(GID) into the radar data processing of narrow band radar for the improvement of narrowband radar data accuracy and the measurement information completion of the target task. The sequential connection of uncontinuous tracks is achieved. Secondly, based on the above, this paper proposes a method of target recognition based on narrow band radar. In this method, the target type and target’s intention are preliminarily judged by the radar track matching algorithm based on correlation factor. In order to achieve the accurate matching of target recognition and to improve the accuracy of narrowband radar target recognition, RCS time series is used to extract the feature. The main content of this paper are summarized as follows:1. In order to utilize data fusion technology in radar network, we firstly establish a reasonable and uniform work platform(coordinate system). In consideration of the effect of earth curvature in radar network which has three radars, this paper firstly converts geodetic coordinate data(latitude, longitude, altitude) of radar station location in three sub-coordinate system to an united terrestrial coordinate system,where coordinate of three radar station location are used to determine one searching virtual plane(work platform). Then through coordinate transformation and data transformation, measured data of three radars will be converted to this virtual plane. Follow-up information processing will be conducted based on this platform. Then obtained coordinate of the target will be converted to terrestrial coordinate system through coordinate transformation. At last, the longitude, latitude and altitude of target will be obtained through geodetic coordinate transformation, which as accurate representation of the position of the target. As the working platform is based on the geodetic coordinate system, the radar target data can be matched with the GID and embedded into any GIS platform. It can effectively express the data information from the sub coordinates, which is conductive to data transmission and integration.2. This paper realizes three-dimensional positioning of the target on the virtual plaatform with the radar networks and solves the problem of height measurement in meter-wave radar that is caused by multi-path effectiveness. This paper improves minimum error method and uses azimuth angle information with not high resolution ratio but good data stability to determine hunting zone of algorithm. The target distance information with high resolution ratio is used to obtain final longitude, latitude and altitude estimate of target. Sometimes target distance estimate may be inaccurate as a result of strong reflection on earth surface, according to which this paper establishes a confidence judgment criterion to verify availability of positioning. Through simulation analysis, this algorithm is verified to obtain a good accuracy in height measurement and can be regarded as an effective method in height measurement for radar network. When the number of useable radar for monostatic or bistatic, two coordinate traditional long-range air defense radar cannot measure height, this paper presents radar detection power auxiliary height finding algorithm combined with GID. The algorithm uses real-time target location of radar detection power contour map to determine the altitude of the target. The result can also be used as the initial value of fine altimetry to provide reliable reference for highly accurate positioning.3. In order to accomplish the task of distance air-defense surveillance and target acquisition, the target echo is often located in low elevation. Because of the radar electromagnetic wave is obscured by the ground object, the target is to be lost in the angle of the radar. Therefore, this paper proposed double threshold blind tracking algorithm which combines GID. The algorithm uses GID to calculate the possible shadowing angle and time of flight target in real time. When the target enters into the shielding angle, this algorithm will use the target tracking information, gets the target’s motion crition, predicte the blind track and set the time threshold and position threshold in the shadow angle. If the target is detected again in a limited time and position threshold, the target is identified as the same as the target covered before. Using the new measurement data to estimate the blind tracking state, the continuity of target tracking is ensured. This method can also be applied to solve the problem of the "track data relay" between the adjacent radars, that is, the problem of the track cross link between the adjacent radars detection blind area. Sometimes in order to maximize the coverage of the surveillance of airspace, the power range between radars is distributed to the whole surveillance airspace in the form of non overlapping. In this case, there will be a small gap between the cover range of the adjacent radar, which is similar to the detection blind area. Based on the work platform of the earth coordinate system, the paper presents the equivalent conversion of the track data of the radar and the track data of the radar. Using double threshold blind tracking algorithm can achieve full spatial track sequential connection. In addition, the remote warning radar will have the uncontinuous track problem caused by target fluctuations.With the influence of the ground reflection path, echo signal-to-noise ratio of the target undulates acutely. That leads to the loss of the target traced by narrow band radar. Combined with MUSIC algorithm, the conventional Toeplitz construction method is modified properly to realize the targets’ signal detection and high-resolution spectrum estimation in velocity-dimension. In order to reduce the impact of multi-path effect, the correlation analysis based on Fourier Transform is adopted. This method can be used to separate the signal eigen subspace and the noise one by the distance relationship of eigen frequency. It can improve detection of targets and velocity resolution in lower SNR case. Both simulation and the actual data testify the validity of this method.4. Comparing the experiential track with the current track detected by radar can identify the aerial target rapidly and make a rough judgment of its navigational intention. Therefore, this paper proposes the algorithm that is used to judge the consistency of radar track based on correlators. The approaches of preset threshold and subdivision approximation are used to determine the experiential track and track points which match the current track points and to work out yaw factor, altitude factor, velocity factor and direction factor. According to the factors mentioned above, the correlation measure criterion of the current track points and the experiential track is constructed in this paper. The K/N norm is used to identify the correlation degree of the current track and the experiential track, and then the judgment of the track consistency is accomplished. The experimental results show that this judgment algorithm, with simple structure and a certain adaptive ability, is easy to implement in engineering. The method can also monitor the relationship between the dynamic tracks and the track of experience to provide real-time early-warning of abnormal tracks. The location of the site can be inferred from the target type with GIS. It can use a priori knowledge of the aircraft type of airport to do target identification.5. In order to solute target recognition problem under commissioned radar system, the paper proposes a target recognition algorithm based on RCS time seriest achieve the accurate recognition of the target type. Firstly, the algorithm uses the track information database formed by long-time observation and radar track match decision algorithm based on correlation functions mentioned above to make real-time track match between dynamic track and experiential track data for the purpose of rapidly finding their RCS time series under the same visual directional angle. In the algorithm, EMD method is applied for the first time to extract the eigen of RCS time series. Then it takes unitary instantaneous frequency of high-frequency eigenfunctions as recognition eigenvalues and sets the valid target recognition criterion. Both simulation and the actual data testify the validity and stability of this algorithm.