ملخص البحث :
Despite many developments and improvements in mobile devices, it still suffers from its limited capabilities such as short battery life, limited processing capabilities, and small storage spaces that can be overcome by canceling the implementation locally and attributing it to the cloud, in this research all cases have been applied to methods Implementation in mobile applications, which can be either Java methods, (C++) methods, or CUDA methods that require calculations to be performed on the cloud server's graphics processor (GPU).We have been working on proposing and developing an algorithm for decision-making in attributing implementation before starting it, according to the network status, as well as developing an application using (Android Studio) accompanying this research, through which the best methods of implementation can be determined and thus it becomes possible to make the final decision in determining a clear mechanism for implementation and choosing between two states, either the implementation is local or on the cloud server. The results are set for each possible implementation case (locally - remotely), based on the decision-making algorithm by determining the implementation time taking into account the mobile device CPU speed, network performance, application program characteristics, and cloud server efficiency. It is important to rely in this research on taking into account the CPU speed of the mobile device, network performance, application program characteristics, and the efficiency of the cloud server to reach the best results that were reviewed in this research with the use of different platforms to choose the best one in accomplishing our work. The actual results of the applied algorithm proved a significant saving in implementation time with the increase in the request to perform tasks on the cloud server and with the increasing complexity of the issue.
-
سنة النشر : 2023
-
تصنيف البحث : scopus
- تحميل