相控阵雷达仿真系统并行计算研究

Parallel Computing in Phased Array Radar System Simulation

  • 摘要: 针对共享存储多处理器的集群环境,研究了减少相控阵雷达仿真系统运行时间的并行计算方法. 根据相控阵雷达多个波束处理过程的相对独立性和信号处理仿真中多个距离单元采样点的计算特点,提出了一种粗粒度消息传递接口(MPI)分布式内存和细粒度OpenMP共享内存混合编程的两级并行方法. 实验结果表明,并行计算使系统的仿真速度有较大提高.

     

    Abstract: Under a symmetric multiprocessors(SMP) cluster environment, parallel computing of phased array radar simulation system is investigated to reduce simulation time. A coarse-grained message passing interface(MPI) distributed memory and fine-grained OpenMP shared memory hybrid parallel computing method is presented based on the relative independence of the phased array radar multibeam processing and the distance unit sampling computing in the signal processing simulation. Experimental results showed that the simulation speed is greatly enhanced by parallel computing.

     

/

返回文章
返回
Baidu
map