نوع المشرف

مشرف رئيسي

تاريخ الاشراف على الرسالة من

2017

الى

2019

اسم الطالب

Rafid Kahtan

ملخص الرسالة

Computer systems have evolved from monolithic computer devices to a much more complex client-server environment in previous years. One of those newly developed technologies is a mobile agent. Because of their innovative capabilities and attractive application, mobile agents have long captured the attention of researchers and industry. Mobile agents are computer programs that can automatically migrate from host to host, transfer their internal state, and enable them to perform tasks more conveniently, more robustly, and more efficiently than traditional client-server applications in network and distributed environments. In this work, the best path in minimal time is found by migrating the mobile agent from the source node to the destination node using mathematical processes and optimization techniques. A genetic Algorithm is used to overcome the problem of selecting the best path rather than the shortest path that is used in previous work. The shortest path is not the mean optimal path at all times. This work focuses on how to minimize the number of nodes that are used to transfer data from source to destination using combining the sequential nodes from a time point view using the Node Compression Algorithm (NCA). When comparing adaptive algorithms with each other, time is considered an essential measure for selecting the best path (minimum time). Using time measurement after 10 optimization iterations as a minimum and holding to the result. The proposed approach shows that using hybrid approaches GA and NCA reduce the time of selecting the best path using GA from 336.448 ms to 286.29 ms for the 10th iteration and reduce the time in other iterations as well. These results show the importance of using optimization techniques in cloud computing to help overcome the distributed nature and minimize the time of transferring data within minimum time complexity. Keywords: Mobile Agent, Genetic algorithm, Push-All-Data Strategy, Design Pattern (Master-Slave), Node Compression Algorithm.