Development of an Accurate Model for Memristor based Temperature Sensor

Memristors are a unique class of electronic devices and have been considered one of the basic circuit elements along with Resistors, Capacitors, and Inductors. A memristor's built-in nonlinear ability is to remember how much charge has passed through it, and memristance is a charge-dependent resistance. This memristance is a variable value that depends on the change in width of the doped region (state variable) of the memristor that in turn depends on various factors like applied voltage, magnetic conditions, temperature, ionic mobility etc. As the temperature changes there is a change not only in the diffusion rate but also the lowest and highest attainable memristance. As there is a change in the memristance at a particular temperature and voltage the write time can be calculated. This research study investigates the possible write time at various temperatures and observe its effect with various window functions. This has been done through analytical modeling of memristor by employing existing non-linear window functions. These changes due to the effect of temperature can be used to model a temperature sensor at the nanoscale level which can significantly reduce the power and area consumption. A detailed study is presented in the current work. The uses of a temperature sensor are very wide and have great effect on exploration of memristor based sensors.