Research On Signal Processing Technology Of Opportunistic Array Radar Under The Uncertain Conditions

The opportunistic array radar(OAR) is a new concept of radar system, which is based on the stealth of the platform and the digital array radar. OAR can perform a varity of functions, such as searching, tracking, fire controlling, guidance, et al. Meanwhile, OAR includes serval working modes. OAR achieves the selection and management of radar resources opportunistically through the perception of the battlefield. Radar waveform design, target detection and the array pattern synthesis are important parts in radar signal processing. Radar waveform design not only determines the signal processing method, but also directly affects many performances, such as the resolution of the radar system, measurement accuracy and the ability of the anti-interference. The constant false alarm rate(CFAR) in radar target dection can avoid changing the detection threshold by the noise, clutter and interference signals. The flexible array pattern can help radar to achieve its functions. But OAR as a complex system, there are some uncertain factors which can affect the overall performance of the radar. These uncertain factors are derived from the changes of the battlefield situation, the failure of the part components, the interference of the internal noise in equipments and et al. Considering the influence of the uncertain factors, the signal processing of OAR needs to consider the uncertainty.Combined with the uncertainty theory, this thesis mainly studies the radar orthogonal waveform design, CFAR algorithms in target detection, array pattern synthesis and the correction of the array error. The main contributions of this dissertation are listed as follows:1. Combined with the characterstics of OAR, such as the distribution of the array elements, the working condition of the elements, tactical functions, operating modes and resource managements, the uncertainty theory which contains the probability theory, credibility theory, opportunity theory is applied to the OAR. The sources of the uncertainty factors in OAR are analysized in detail, and the signal processing theory and method for OAR is studided.2. Design the orthogonal polyphase waveform and discrete frequency coding waveform(DFCW) with better autocorrelation and cross-correlation. Based on the analysis of the autocorrelation and cross-correlation properties, combined with the genetic algorithm(GA) and simulated annealing algorithm(SA), considered the gray uncertain factors between the autocorrelation and cross-correlation properties, introduced the gray correlation evaluation to analyze the fitness function, the hybrid genetic simulated annealing algortih(HGSAA) is proposed to design orthogonal waveforms. The simulation results show that the new algorithm HGSAA can design the orthogonal signals with better properties.3. Propose the distributed multi-sensor CFAR with the maximum censored mean detector(MX-CMLD) CFAR based on the fuzzy logic. On the basis of the common CFAR which is based on the probability theory, combined with the credibility theory, the common binary detection with hard discriminant 0 or 1 is instead of the soft discriminant through the membership function. That is to say, the membership function maps the observation space to a value between 0 and 1 to represent “no signal” and “signal” hypothesis. Combined with greatest of(GO) CFAR and censored mean level detector(CMLD) CFAR, the distributed fuzzy MX-CMLD-CFAR based on the fuzzy logic is proposed, and a variety of fusion rules is deduced. Finally with the fuzzy algebra product rule as example, the distributed fuzzy MX-CMLD-CFAR is studied. The simulation results show that the distributed detection system not only does a better job of maintaining CFAR in homogeneous background, in the presence of interfering targets and in the clutter boundary, but also is superior to binary distributed detection system. At the same time, the distributed fuzzy system can avoid the effection of the part sensor failure.4. Propose the distributed multi-sensor CFAR with credibility degree based on the voting fuzzy fusion. Considering the random distribution characteristics of the array elements and uncertain working modes, the credibility degree of each sensor should be considered in the fusion center. On the basis of the fuzzy MX-CMLD-CFAR, design the credibility function with expert experience to describe the credibility degree of each sensor. In the fusion center, the local membership function of each sensor are fuzed to produce the global membership function via voting fusion according the credibility of each sensor. The simulation results indicate that the distributed fuzzy CFAR can perform better than common binary detection and avoid the effection of the part sensor failure.5. Propose a new pattern synthesis algorithm based on the fuzzy chance-constrained programming and fuzzy dependent-chance programming for antennas in a single area. Considering the antenna elements with random distribution and excited state of uncertainty, using fuzzy variables to characterize the complex and uncertain environment of antenna elements, create the programming model for different constraint conditions and objective functions. The fuzzy chance-contrained programming, which based on the number of antenna elements as constraint conditions, can minimize the sidelobe level and the error of the main lobe bandwidth under a certain credibility degree. The fuzzy dependent-chance programming, which based on the number of antenna elements as the constraint conditions, let the chance of sidelobe level meeting conditions maximize. This model can solve some disadvantages of the fuzzy chance constrained programming model, such as the final result can’t achieve the origin optimization results under the constraint conditions.6. Propose a new pattern synthesis algorithm based on fuzzy random dependent-chance programming for antennas in different areas. Considering the OAR need to synthesize the different pattern for multiple radar tasks at the same time, the total number of antennas for pattern synthesis is limited by the radar resources. The number of antnennas for parttern in each region is regarded as the fuzzy random variable. With constraints, create the fuzzy random dependent-chance model to make the chance of patterns meeting the conditions maxsize.7. Propose the pattern synthesis optimization with mutual coupling based on the sub-array. Considering the big number of antenna elements and the random distribution, the large array is divided into serval sub-arrays according to the special distance. Calculate the generalized mutual impendance matrix of sub-array with method of moment to build the mutual impendance matrix of the large array. Use adaptive pattern synthesis or genetic algorithm to implement the pattern amendment with mutual coupling. The simulation results show that the amendment algorithm based on the sub-array can reduce the influence of mutual coupling.