Study of logic synthesis and optimization

Abstract

With larger and larger scale of integration the problem being faced by the technology is that logic circuitry are becoming more and more complicated. It is important to be able to optimize logic circuitry, so that circuitry becomes as simple as possible. In this thesis a set of important logic synthesis and optimization methods have been thoroughly studied theoretically, and relative performances of the three of them have been evaluated by carefully designed experiments. In these experiments we apply different types of functions and observe the time elapsed, number ofliterals, and number of products. For the same set of data we have applied different orientation of data. This study shows that for Quine- McCluskey and M1ynarovic methods both the number of products and the literals are the same but for EXMIN2 the number of products and the number of literals differs in a smaIl extent. When the number of variables increases all the methods need more time, but in the case of M1ynarovic method the time elapsed is fixed for a particular number of variables whatever the number of inputs may be. Time required for Quine-McCluskey and EXMIN2 varies with number of inputs for the same number of variables. For the same set of data EXMIN2 gives the minimal number of product terms for the Arithmetic and Symmetric function. For randomly generated function QIlino-McCluskeymethod gives fewer number of products.than M1ynarovic method but number of literals is less in most of the cases in the later inethod. When. number of inputs is less comparable than the maximum possible inputs then Quine-McCluskey method gives fewer number of products than M1ynarovic method. It is seen that from the point of view of product terms and literals EXMlN2 is better than the others. The thesis also gives an elaborate guideline for future research work to proceed in this direction.