Self-Learning Analog Comparator with Adaptive Sampling Rate Scheme for Energy Optimization in Continuous Input Monitoring Applications.

Abstract

The analog comparator module integrated on the microcontrollers is used in several embedded applications in which analog input signals from sensors and other sources need to be monitored continuously. These applications require consuming very low energy as they operate from battery or powered through energy harvesting sources. This implies that every module integrated on the microcontroller must be designed for lowest energy consumption. There are different implementation and operating techniques followed for low energy consumption of analog comparator.

This paper describes an operational scheme in which the analog comparator module learns the nature of the analog input signal from the previous comparison results and dynamically adapts the sampling rate for the lowest energy consumption without compromising the functional performance. The explained scheme is completely transparent to software and enables intelligent hardware operation through self-learning technique.