Lifetime improvement using task scheduling in hard real-time systems equipped with environmental energy harvester and hybrid energy storage

In this paper, we investigate the effect of task scheduling on the lifetime of a real-time hard drive system that uses a composite energy tank consisting of a battery, a supercapacitor, and a solar energy picker to power supply itself. The life of a system is defined in terms of the moment the system starts up to the moment its tasks are disrupted due to lack of energy. Due to the nonlinear properties of batteries and supercapacitors, which cause their internal charge to be divided into available (IAC) and inaccessible (IUC), the lifetime of such a system depends entirely on the charging and discharging pattern of the energy tank, as this pattern ultimately leads to the amount of charge stored in the IUC section and the amount of charge extracted from this section. Therefore, we can influence the lifetime of the system and increase it by managing the charge / discharge pattern of energy tanks. Since the pattern of energy delivery from the environment is beyond the control of the system, we try to influence the tank charge / discharge pattern by adjusting the consumption pattern to ultimately improve the life of the system. In this regard, we have presented two scheduling algorithms MCF and MGF, which are respectively trying to perform the most consumed and least consumed task in the system, and then using the MCG policy, which at any time according to the conditions for using one of these decision algorithms are presented. Experimental results show that we can increase system lifetime by between 5% and 16%. Considering that in recent years the issue of using supercapacitors along with batteries and solar cells in space systems has been raised, so the results of this research can be investigated for use in satellites.