A New Method for a Fractal-Type Temperature Sensor Using Cellular Oscillatory Networks

Abstract

A new fractal temperature sensor structure using simple cellular oscillatory networks is proposed. The frequency of the cellular oscillatory networks is analyzed in terms of threshold voltage and mobility, and then expressed as a function of temperature. In order to validate this new method, sensor networks with 405 inverters is used in the network, but it can be expanded infinitely in two-dimensional or even three-dimensional fractal structure. Through simulations, we found that local temperature change yields an instant global frequency drift in the fractal structure. For analysis, we used VLSI simulations with specifically BSIM4 model parameters (level54 using HSPICE). Compared with other temperature sensors, this structure significantly increased operating sensing range, − 50 °C to + 150 °C, with maximum 6 μW power consumption. With standard 100 nm CMOS MOSFET devices, less than ± 1.8 °C (0.9%) sensing accuracy, with 1 V supply voltage was achieved.