Affiliation:
1. PEC University of Technology, India
Abstract
Digital signal processing algorithms are recursive in nature. These algorithms are explained by iterative data-flow graphs where nodes represent computations and edges represent communications. For all data-flow graphs, time taken to achieve output from the applied input is referred as iteration bound. In this chapter, two algorithms are used for computing the iteration bound i.e. Longest Path Matrix (LPM) and Minimum Cycle Mean (MCM). The iteration bound of single-rate data-flow graph (SRDFG) can be determined by considering the Multi-rate data-flow graph (MRDFG) equivalent of the SRDFG. SRDFG contain more nodes and edges as compared to MRDFG. Reduction of nodes and edges in MRDFG is used for faster determination of the iteration bound.