Open AccessANDROID SOFTWARE AGING AND REJUVENATION MODEL CONSIDERING THE BATTERY CHARGE
Author(s) -
Vitaliy Yakovyna,
Bohdan Uhrynovskyi
Publication year - 2022
Publication title -
radìoelektronìka, ìnformatika, upravlìnnâ/radìoelektronika, ìnformatika, upravlìnnâ
Language(s) - English
Resource type - Journals
eISSN - 2313-688X
pISSN - 1607-3274
DOI - 10.15588/1607-3274-2021-4-13
Subject(s) - rejuvenation , state of charge , android (operating system) , computer science , software , markov chain , battery (electricity) , markov process , simulation , reliability engineering , engineering , operating system , machine learning , gerontology , medicine , power (physics) , statistics , physics , mathematics , quantum mechanics
Context. A feature of mobile systems is their dependence on battery charge, which is an important factor when planning various processes, in particular when planning time of performing software rejuvenation procedure.
Objective. The goal of this article is to develop a model of software aging process with performing rejuvenation procedure for the Android operating system considering the factor of battery charge.
Method. A complex model based on Continuous-Time Markov Chains is proposed, which combines the software aging and rejuvenation model, the user behavior model and consider battery charge factor. A graph of states and transitions describing a complex model is constructed. Based on the formed graph the system of differential equations is written. The system was calculated using the 4th order Runge-Kutta method. The optimal time for the rejuvenation procedure can be determined when rejuvenation will not interfere with the user and will be performed before the battery is fully discharged, ie when the probability of the system being in these states is the lowest.
Results. The simulation of the developed model for test values of transition rates is performed. Considering the battery charge model allows to avoid planning the rejuvenation procedure at a time when the mobile device is likely to have a low charge or be completely discharged.
Conclusions. The proposed model based on the Markov chain allows to predict the start time of software rejuvenation procedure, considering both user behavior and battery level, which can have a significant impact on the predicted time. Also, the early implementation of the rejuvenation procedure may have the effect of reducing the system workload and delaying the discharge of the device, which should be checked in further studies. The expediency and importance of the consideration of battery charge factor and the need for further study of the proposed software aging and rejuvenation model are substantiated.