DSSA: DIRECT SIMPLIFIED SYMBOLIC ANALYSIS USING METAHEURISTIC-DRIVEN CIRCUIT MODELLING

Symbolic simplification of analog circuits is a computationally challenging problem due to the exponential growth of the number of sym-bolic terms with the circuit size. In recent years, researchers have proposed various methods to address this problem, but these methods often require matrix-or graph-based symbolic analysis methods, which can be computation-ally expensive and memory-intensive, especially for real-size analog circuits. To overcome these limitations, we introduce a new methodology called Di-rect Simplified Symbolic analysis (DSSA). The proposed DSSA method views simplified symbolic circuit analysis as a modeling problem and directly pro-duces the most significant terms of the transfer function, without the need for traditional circuit analysis. One of the main advantages of DSSA is that it sig-nificantly reduces computational complexity and required memory compared to the existing techniques. This is achieved by generating a dataset using the Monte-Carlo simulation method and performing a genetic algorithm to solve the established modeling problem. The objective is to minimize the average numerical error between the simplified symbolic expression and the exact nu-meric expression for all data points. The proposed method has been tested on five circuits in MATLAB, and the results clearly demonstrate its perfor-mance against existing methods. The findings by the DSSA algorithm across five circuits reveal 0.64 dB and 1.36 dB variations for the average and maxi-mum dc-gain, respectively. Moreover, the DSSA algorithm exhibits an average pole/zero error of 6.8% and a maximum pole/zero displacement of 16.8%. It has the potential to improve the efficiency and accuracy of symbolic analysis, making it a promising tool for circuit designers and engineers.