COMPARATIVE STUDY OF ULTRA-LOW-POWER MICROCONTROLLERS IN ENERGY-AUTONOMOUS ENVIRONMENTAL WIRELESS SENSOR NODES (WSN)

Climate change poses great danger for forest ecosystems by impacting the carbon cycle at a global and regional scale [1]. By monitoring its effects using distributed environmental sensor systems, a deeper understanding of cause-effect chains can be obtained, as a basis for the development of damage control strategies. For reliable and maintenance free operation, these sensor nodes should be ultra-low-power systems, supplied either by long-lasting batteries or by energy harvesting. Here, one deciding factor for energy optimization is the microcontroller choice which depends on the complexity of the required application, the required computational power and the resulting energy consumption. This choice is difficult as there is often a tradeoff between processing capabilities and energy consumption. In this contribution an energy analysis for different microcontroller platforms with architectural differences, namely the MSP430FR5994 and the ARM Cortex M4F-based Ambiq Apollo 3 is provided. Their processing capabilities are exploited while performing data acquisition and complex mathematical computations. The resulting energy consumption analysis shows that the MSP430FR5994 is suitable for lower complexity wireless sensor nodes (WSNs) whereas the Apollo 3 is better suited for high-end computational WSNs. Thus, this contribution serves as a guide to identify the most suitable microcontroller for a certain environmental WSN.