An efficient method to compute simplified noise parameters of analog amplifiers using symbolic and evolutionary approach

摘要

Symbolic circuit analysis suffers as the symbolic terms increases by many folds with the increase in the size of the circuit. Hence simplification techniques are used for the analysis of the circuits which are larger in size. In this work, multi-objective evolutionary optimisation algorithms for obtaining simplified symbolic expressions of analog amplifier circuits are presented which are not only very efficient in terms of speed and the amount of circuit reduction but also solve the accuracy problems. Initially, all MOSFETs are replaced with their nullor equivalent model. Then, the formulation of accurate fully-symbolic noise expressions is carried out by the application of symbolic nodal analysis. The derived exact symbolic expressions are then simplified using an evolutionary approach. This method has been successfully coded in MATLAB. The numerical results obtained before and after simplification are compared with SPICE simulation to demonstrate the efficiency of the proposed methodology.