Experimental research of passive bistatic radar based on pipeline processing

For passive bistatic radar (PBR), it is convenient to use software method to realise the signal processing. However, because of the huge amount of data computing, it is difficult to achieve the real-time signal processing by software method only with server that uses CPU for calculation. To solve the problem, the PBR signal processing method with multi-CPUs based on pipeline processing is studied. The key point of the method is to improve the parallelism of the signal processing algorithm and to use all the CPU cores efficiently for parallel computing. The whole PBR signal processing is divided into several steps, such as clutter cancellation, correlation processing, CFAR detection and so on, and each step employs different CPU resources that contain many cores. The data processed in each steps is also divided into several pieces to fully use all the CPU cores. At last, the signal processing in software for digital television terrestrial broadcasting (DTTB) signal is achieved based on the method. The experiment is carried out to verify the method. In the experiment, the whole signal processing time for DTTB signal is much less than the time of signal to be processed, and the signal processing output is continuous.