Artificial Neural Networks-based Ambient RF Energy Harvesting with Environment Detection

This paper proposes a cascading artificial neural networks (ANNs) algorithm with performance-enhancing filters for ambient radio frequency (RF) energy harvesting (EH) with environment detection (ED), referred to as ANN-ED. This ANN-ED algorithm can reliably operate in both urban and rural environments where there is unpredictable availability of unintended sources and dynamic channel conditions between the sensors and the unintended sources. Numerical results show that sensors using the ANN-ED algorithm can successfully sense up to 98.7% of the data compared to an ideal sensor, offering a significant improvement compared to the 0.3% achieved by an ANNs-based RF EH without ED. Sensors using the ANN-ED algorithm have an accuracy rate of up to 100% as well; a significant improvement over that of an ANNs-based RF EH without ED whose accuracy can be as low as 0%. The reliable operation of ambient RF EH sensors in all environments enhances the practicality of its usage regardless of location.