基于GPU的天基预警雷达信号自适应检测仿真

Adaptive Detection Simulation of Space-Based Early Warning Radar Signal Based on GPU

针对天基预警雷达信号检测过程中信号和噪声特性未知的问题,提出了一种基于GPU的天基预警雷达信号自适应检测方法。首先,使用纹理类型处理GPU储存器,使得后期的信号检测更为简便,使用多向访问技术和对称计算方法计算处理后的GPU,得出GPU储存器的信号数据访问次数,依据访问次数,设计出GPU处理核心,从而更为简便的寻址,构建参数可调的非线性SR装置,在不同频率的雷达信号下,通过SR自适应检测对其地址中的频段进行自适应检测,并依据不同情况重新设定系统参数和频段范围重复检测,直至GPU的输出理想,完成天基预警雷达信号自适应检测。仿真结果证明,检测天机预警雷达信号过程中,检测信号输出较为完整,且输出功率趋于稳定。

In this article, an adaptive detection method for space-based early warning radar signal based on GPU was proposed. Firstly, the texture type was used to improve GPU memory, which made the signal detection easier in later stage. Then, the multi-direction access technology and symmetrical computation were used to calculate the GPU after processing, so that the number of accesses to the signal data of GPU memory was obtained. According to the number of accesses, GPU processing core was designed, so that the addressing became more and more convenient. Furthermore, non-linear SR device with adjustable parameters was constructed. Based on the radar signals with different frequencies, the frequency band in its address was adaptively detected by SR adaptive detection. Due to different conditions, it was necessary to reset system parameters and frequency and thus to carry out the repeated detection until the output of GPU was ideal. Finally, the adaptive detection of space-based early warning radar signals was completed. Simulation results show that the signal output is relatively complete and the output power tends to be stable in the process of detecting the signal of early warning radar.