ADROIT: Adaptive Recharge Optimization in Urban IoT

Urban IoT facilitates the automatic and remote collection of environmental and societal data streams with detailed spatial and temporal resolution. However, energy-constrained IoT nodes face challenges due to frequent recharging, which is both resource-intensive and disrupts data continuity, impacting continuous applications like environmental monitoring and traffic management. Existing approaches often fail to handle the dynamic nature of urban sensing characterized by heterogeneous data significance. In an attempt to address this issue, we put forth an adaptive approach, ADROIT, to adjust the energy level (EL) of IoTs, inspired by the concept of electron ELs in quantum systems, where the sensing parameters are modulated to preserve event data distribution while minimizing recharges. Specifically, a coordinating node adjusts the EL of each IoT based on their local neighborhoods. Our preliminary analysis shows that ADROIT adapts the ELs based on neighboring nodes' activity. By integrating dynamic similarity measures and policy gradient algorithms, ADROIT promises enhanced adaptability to event frequency and importance as well as spatial distribution.