Sie befinden Sich nicht im Netzwerk der Universität Paderborn. Der Zugriff auf elektronische Ressourcen ist gegebenenfalls nur via VPN oder Shibboleth (DFN-AAI) möglich. mehr Informationen...
Optimising operational amplifiers by evolutionary algorithms and gm/Id method
Ist Teil von
International journal of electronics, 2016-10, Vol.103 (10), p.1665-1684
Ort / Verlag
Abingdon: Taylor & Francis
Erscheinungsjahr
2016
Quelle
Taylor & Francis
Beschreibungen/Notizen
The evolutionary algorithm called non-dominated sorting genetic algorithm (NSGA-II) is applied herein in the optimisation of operational transconductance amplifiers. NSGA-II is accelerated by applying the g
m
/I
d
method to estimate reduced search spaces associated to widths (W) and lengths (L) of the metal-oxide-semiconductor field-effect-transistor (MOSFETs), and to guarantee their appropriate bias levels conditions. In addition, we introduce an integer encoding for the W/L sizes of the MOSFETs to avoid a post-processing step for rounding-off their values to be multiples of the integrated circuit fabrication technology. Finally, from the feasible solutions generated by NSGA-II, we introduce a second optimisation stage to guarantee that the final feasible W/L sizes solutions support process, voltage and temperature (PVT) variations. The optimisation results lead us to conclude that the g
m
/I
d
method and integer encoding are quite useful to accelerate the convergence of the evolutionary algorithm NSGA-II, while the second optimisation stage guarantees robustness of the feasible solutions to PVT variations.