Implementation of software-based sensor linearization algorithms on low-cost microcontrollers.

Nonlinear sensors and microcontrollers are used in many embedded system designs. As the input-output characteristic of most sensors is nonlinear in nature, obtaining data from a nonlinear sensor by using an integer microcontroller has always been a design challenge. This paper discusses the implementation of six software-based sensor linearization algorithms for low-cost microcontrollers. The comparative study of the linearization algorithms is performed by using a nonlinear optical distance-measuring sensor. The performance of the algorithms is examined with respect to memory space usage, linearization accuracy and algorithm execution time. The implementation and comparison results can be used for selection of a linearization algorithm based on the sensor transfer function, expected linearization accuracy and microcontroller capacity.