Study On Wideband Digital Beam Forming For Multi-Beam At Subarray Level

Digital beam forming is one of the hotspot technology which can dramatically increase the performances of the modern radar, electron reconnaissance and wireless communication system. The application of DBF is limited by the high cost of digital receiver/transmitter elements. Modern Radar systems usually adopt the form of subarray in order to reduce the number of digital channels. The wide band multi-beam forming technology at uniform subarray of large array is discussed in this paper.1 The two typical thoughts of designing constant beam-width beam-former are researched, and the method called spatial resample based on sampling theorem is analyzed and designed both in frequency and time domains. Simulation shows that this method is simple and easy to implement.2 The algorithms of multi-beam forming both for linear array and planar array include separate weight multi-beam former, FFT multi-beam former and multi-linearly constraint minimum variance multi-beam former are researched and simulated, and their advantages and disadvantages are compared.3 The performance of multi-beam at uniform subarray level is researched, two method of multi-beam former at uniform subarray level are proposed for the purpose of reducing the grating lobe and increase the performance of multi-beam.A method called spatial interpolation based on lag estimation is developed for multi-beam forming at uniform subarray. Due to the more than half wavelength distance between each subarray the grating lobes appear, this method insert values into every half wavelength position based on the lag estimation of adjacent subarray to suppress the grating lobes, simulation proves this method is effective and limitative.Because of the limitation of spatial interpolation multi-beam former, a method based on genetic algorithm is developed. This method is used to suppress the grating lobe and increase the performance of multi-beam through designing the weights of multi-beam at elements level and subarray level, and each weight is optimized by genetic algorithm. Simulation shows that this method can increase the performance of line array and plan array at different degrees. For the purpose of solving the low real time problem of GA, a proposal of optimizing the weights offline, saving, and then loading these weights when the system work is discussed.4 The algorithms of adaptive digital beam forming is discussed, and take example for SMI algorithm, the SMI algorithm at subarray level is researched.