CMOS Computational Structures Using a Nonlinear Multifunctional Core

Signal processing circuits represent an important class of CMOS structures, finding a multitude of applications: neural networks, portable equipments, wireless nanosensors, communications and medical implantable devices. The analog implementation of these circuits can improve their performances, the low-power operation, reduced silicon area and increased processing speed representing the most important advantages. We will present a new approach of designing analog signal processing circuits, introducing the new concept of multifunctional structures (computational circuits that use the same functional core in order to obtain multiple linear and nonlinear circuit functions). The efficiency of the proposed design technique will be illustrated for obtaining high performances fundamental analog signal processing circuits: transconductance amplifiers, voltage multipliers, squaring blocks, active resistor structures and function synthesizer circuits. The estimated theoretical results confirm by specific simulations.