基于GPU的高动态导航信号实时仿真生成

High Dynamic Navigation Signal Real Time Simulation and Generation based on GPU

针对软件仿真高动态扩频信号实时性差的问题,通过分析扩频信号生成过程和分段插值运动模型,建立了一种高动态信号生成模型,由动态参数和当前时间计算瞬时伪码相位和瞬时载波相位,各采样点的计算彼此不相关,将仿真信号组织成多维的数据结构,利用GPU中大量的浮点运算单元并行计算,提高了信号仿真的实时性。仿真验证结果表明,基于GPU的分段插值多维并行算法仿真时长1s采样率100Msps的信号,仿真与传输耗时共计322ms,与CPU串行仿真相比加速约19.4倍,实现了高动态扩频信号实时仿真。

For solving the real - time problem of high dynamic spread spectrum signal simulation, a new high - dynamic signal generation model is proposed. The instantaneous pseudo - code phase and instan- taneous carrier phase calculate from the current time and the dynamic parameters. Calculation for each sam- pling point is not related to each other, so the simulated signal sampling points are organized into multidi- mensional data structure. By using parallel computing with a large number of ALU in GPU, the simulation speed is improved. Simulation results show that, with the multidimensional parallel interpolation simulation algorithms based on GPU, simulation and transmission consuming 322ms in all for 100Msps per second sam- pling rate signals. Compared with the CPU simulation,the method speeds up to 19. 4 times, high -dynamic spread spectrum signals can be simulated in real time with this method.