Support Vector Machines for Scheduled Harvesting of Wi-Fi Signals

This letter investigates the incorporation of support vector machines (SVMs) to predict the time and location of maximum Wi-Fi coverage for energy harvesting. Such incorporation of machine learning, along with radio frequency energy harvesting systems, improves the proposed rectenna efficiency especially when scavenging wireless routers and access points. Experiments showed that the proposed framework that uses SVM was able to predict the time and location of maximum Wi-Fi coverage. Such prediction constitutes the basis of a scheduling mechanism for directed harvesting of Wi-Fi signals.