Digital Beamforming Algorithm Research And Pc Cluster Parallel Implementation

As an important branch of array signal processing technology, adaptive digital beamforming has been widely used in radar, wireless communication, medical imaging, geological exploration, radio astronomy and so on. Traditional digital beamforming algorithm in theory has good interference suppression ability. In practice, its performance degrades seriously due to the presence of various errors; therefore it is necessary to do some research to the robust beamforming algorithm. MIMO technology is regarded as the key technology to the MIMO radar system, in which the beamforming problem is worthy of further exploring and studing. With the increase of computational burden of adaptive digital beamforming algorithm, organizing the algorithm to improve the speed and efficience becomes a major problem; the PC cluster technology has been widely used in weather monitoring, simulation and other areas of marine with its low cost, easy to build, easy to transplant, etc. But it rarely used in beamforming field, so we may expect to use PC cluster technology to calculate the weight of beamforming in large-scale array.In this work we first explain the basic model of array signal under the narrow assumptions and review the classical beamforming algorithm. An iterative LCMV algorithm was derived on the basis of LCMV algorithm and simulation result verifies its speed and effectiveness. By analyzing the impact of pointing error to the classical beamforming algorithm, we show the necessity of robust beamforming.We then study the robust beamforming algorithm based on convex optimization theory intensively aiming at the drawback of traditional robust beamforming algorithm, including robust Capon algorithm (RCB), iterative robust Capon algorithm (IRCB), the worst-case performance optimization algorithm (WCPO) and the Sequential Quadratic Programming beamforming algorithm based on orthogonal decomposition of the error vector (SQP). These algorithms can show good robustness with pointing error in the desired signal, not only make the pattern closer to the ideal pattern, but also improve the output Signal to Interfere Noise Ratio(SINR). To overcome the difficulty estimate the uncertainty in RCB and WCPO algorithm, we put forward IRCB and SQP algorithm. The IRCB algorithm makes the beam point to the true DOA gradually through iterations, while the SQP algorithm uses the orthogonal components to fix the error steering vector by orthogonal decomposition of mismatch errors. This paper elaborates on the performance of several algorithms and conducted a comparative analysis.The thesis subsequently studies the MIMO beamforming algorithm based on thin array. We first study the thin array optimization method based on standard particle swarm optimization and put forward a new method based on adaptive particle swarm optimization, and then study the MIMO beamforming, in which two MIMO beamforming algorithms are given. Finally, we apply thin array optimization technology to the MIMO array which extends the MIMO array aperture, thereby enhances the MIMO beam resolution.At last we establishe the PC cluster parallel computing systems based on MPI mechanism and tests the system performance. Base on the system we design and implement the software test and management platform on PC cluster, and then we implement the parallel beamforming algorithm previously studied. We describe the implementation process of beamforming algorithm in detail and give the corresponding parallel algorithms which provide an important reference for future research and application.