Periodically Activating and Sleeping Devices in Internet of Things

How to effectively utilize the limited battery capacities is crucial for IoT (Internet of Things) devices. In many applications, it is not necessary to keep every device always active. In other words, devices should be periodically activated and slept to reduce energy consumptions. In this paper, the problem under study with device active time minimization is mathematically modelled using ILP (Integer Linear Programming). After analyzing the bounds of the variables, the IILP (Improved Integer Linear Programming) algorithm is proposed to solve the considered optimization problem in polynomial time whereas it cannot guarantee to obtain the optimal solution. Moreover, the traverse-based TP (Two Pointers) algorithm is developed to obtain the optimal solution with much longer computation time. By comparing IILP to TP over a lot of instances, experimental results show that IILP is much faster than TP whereas TP outperforms IILP in effectiveness.