Model-based design for self-sustainable sensor nodes

Long-term and maintenance-free operation is a critical feature for large-scale deployed battery-operated sensor nodes. Energy harvesting (EH) the most promising technology to overcome energy bottleneck of today’s sensors enable vision perpetual operation. However, relying on fluctuating environmental requires an application-specific analysis statistics combined with in-depth characterization circuits algorithms, making design verification complex. This article presents model-based (MBD) approach EH-enabled devices accounting dynamic behavior components in power generation, conversion, storage, discharge paths. The extension existing compact models data-driven statistical modeling allows accurate offline analysis, verification, validation. presented facilitates optimization during development phase reliable long-term evaluation datasets. Experimental results demonstrate accuracy flexibility this approach: model solar-powered wireless node shows determination coefficient (R2) 0.992, resulting error only -1.57 % between measurement simulation. Compared state-of-practice methods, MBD attains reduction estimated state-of-charge up 10.2 real-world scenario. offers non-trivial insights choices: storage element selection reveals 2–3 times too high self-discharge per capacity ratio supercapacitors peak current constrain lithium-ion polymer batteries.