Abstract
The Directed Search Algorithm performs Iwo-level minimization of switching functions. The identification of prime implicants and the selection of a minimal cover are performed in a synergetic fashion. Purely cyclic situations are resolved by either a minimum method or an approximate method--as specified by the user. The algorithm employs an algebraic topological perspective of a switching function along with viewing the problem as one in Artificial Intelligence involving a state-space search. A search for prime implicants is only performed for minterms which must be covered, with immediate identification of essential prime implicants. If essentiality is not detected, a recursive search is employed using information gained by previous search to reduce the amount of subsequent search. A concept known as pseudo-essentiality is made possible by use of another concept known as the true form cell; these two concepts allow selection of prime implicants sooner than in other methods of switching function minimization. In addition, the true form cell concept allows removal front consideration of superfluous prime implicants. A switching function is partitioned into as many smaller independent problems as possible by use of the concept of the cyclic chain; the cyclic chain is further partitioned by the use of pseudo-essentiality. Only the prime implicants involved in a cyclic chain are considered together for cyclic resolution. The Directed Search Algorithm is implemented in the FORTRAN language.
McKinney, Melvin Howard (1974). A directed search algorithm for the canonical minimization of switching functions. Doctoral dissertation, Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -175802.