摘要
为了提高逆合成孔径雷达(ISAR)实时成像的性能,本文首先设计了一种基于TMS320C6678多核信号处理器(DSP)的高速实时信号处理平台,优化了功耗的同时提高了信号处理能力。其次,本文提出了一种利用窄带信息进行成像预处理的成像流程,该方法利用窄带信息进行目标运动特性分析,并用分析结果指导成像条件判断、成像数据选择和高速运动补偿。根据提出的成像流程将该成像任务分割成几个独立的任务,在分析任务的实时性和任务间的通信的基础上,完成了任务在多核DSP上的分配。利用本文平台对实测数据进行处理,并将成像性能和实时性与单核DSP信号处理平台做对比,进一步验证了多核信号处理平台的处理优势和算法设计的合理性。
Aiming at improving the performance of Inverse Synthetic Aperture Radar（ISAR） real-time imaging, a high speed real-time signal processing platform based on TMS320C6678 Multicore Digital Signal Processor（DSP） is proposed to improve the capability of signal processing besides optimizing the power consumption. Moreover, this paper presents a real- time imaging flow in which narrow-band information has been used to analysis characteristics of target＇ s motion feature. And the analysis results are used to decide image-forming condition, select imaging data and guide high velocity signal com- pensation. We decompose the imaging task into several independent tasks based on proposed ISAR real-time imaging flow, and distribute them among multi cores of digital signal processor on developed signal processing platform according to the time cost of tasks and communications between tasks. Experiment results evaluated on real data verify the superiority of multicore signal processing platform and effectiveness of algorithm design by comparing the imaging and real-time perform- ance to the results obtained by signal processing platform with single-core.
出处
《信号处理》
CSCD
北大核心
2013年第9期1238-1243,共6页
Journal of Signal Processing
关键词
逆合成孔径成像
数字信号处理器
多核
并行处理
ISAR Imaging
digital signal processor （DSP）
Muhicore
parallel processing
