中文摘要：

在分析各种空时自适应处理（STAP）算法特点的基础上，提出了一种用于STAP的并行计算模型．进而深入分析了STAP系统的实时特性，分析发现，要减少系统延时，一是必须尽量减少计算或者通信时间，二是充分利用前批数据的计算结果和相关先验知识；而要增大吞吐率，一是增加处理器流水线条数，二是避免出现瓶颈阶段．基于该模型，已成功开发了多种实际系统．

英文摘要：

The characteristic of space-time adaptive processing （STAP） algorithms is analyzed, and a parallel computation model for STAP is proposed. Then the real time performance is analyzed and some ways to increase the real time performance are found. The effective ways to reduce latency are reducing the latency of different phase independently or using the output of previous pipeline. And the ways to increase throughput are adding more pipeline or avoiding the bottleneck phase. Based on the model, several kinds of STAP systems have been developed.