Graph Based Systematic Synthesis Procedure of gm-C Filter

Real world signals, being analog by nature requires analog signal processing. But unfortunately analog domain still lacks of good research on analog EDA tool. Therefore, developing systematic synthesis methodology is a major research topic still to be addressed and improved to cope up with the recent technological advancements in VLSI. In this background, a graph has been employed to represent synthesis methodology for linear analog filters in a systematic way. Nodes of bipartite graph represents some basic first order mathematical functions; those are connected through weighted edges (with relation operator). Traversal of the graph with defined traversal rules gives different subgraphs implementing different types/order of filters. The graph has been optimized with some defined set of rules. Optimized graph has been converted into a multifunctional filter structure, which realizes all types of filters by programming the connections among the building blocks. This systematic synthesis approach supports reusability and programmability in hardware, when implemented in gm-C. Filter response is verified through SPICE simulation.