An Efficient Method to Threshold Logic Functions Identification

In this paper, a novel method to identify threshold logic functions (TLF) is proposed. Threshold logic is an alternative to conventional Boolean logic, and it is being revisited due to its suitability in emerging technologies, such as QCA, RTD, SET, TPL and Spintronics. Identification and synthesis of TLF are fundamental steps for the development of circuit design flow based on threshold logic. The proposed method exploits both the order of Chow parameters and the system of inequalities to assign the variable weights and the threshold value. This is the first algorithm that not uses integer linear programming (ILP) and is still able to identify all threshold functions up to five variables. Moreover, it also identifies more functions than other methods when the number of variables is higher than five. The average execution time is less than 1 ms per function, similarly to the state-ofthe-art approaches, and the proposed algorithm is scalable. Furthermore, the method always assigns the minimum weights, which results in circuits with less area overhead.