Based on the theories of task interlock and job queue, an efficient method for constructing macropipeline parallel algorithms is presented, and a set of linear and nonlinear macropipeline parallel algorithms for data signal processing are designed. 本文基于任务相关、作业排队等理论,提出构造宏流水线并行算法的一种有效方法,并且利用这种方法构成了一组用于数字信号处理的线性和非线性宏流水线并行算法。