Reducing the Power Consumption of IoT with Task-Oriented Programming

Limiting the energy consumption of IoT nodes is a hot topic in green computing. For battery-powered devices this necessity is obvious, but the enormous growth of the number of IoT nodes makes energy efficiency important for every node in the IoT. In this paper, we show how we can automatically compute execution intervals for our task-oriented programs for the IoT. These intervals offer the possibility to save energy by bringing the microprocessor driving the IoT node into a low-power sleep mode until the task need to be executed. Furthermore, they offer an elegant way to add interrupts to the system. We do allow an arbitrary number of tasks on the IoT nodes and achieve significant reductions of the energy consumption by bringing the microprocessor in sleep mode as much as possible. We have seen energy reductions of an order of magnitude without imposing any constraints on the tasks to be executed on the IoT nodes.