Search In this Thesis
   Search In this Thesis  
العنوان
Modeling of a distributed system For network /
المؤلف
Said, Abdullah El-Araby Mosad.
هيئة الاعداد
باحث / عبدالله العربي مسعد سعيد
مشرف / مجدى زكريا رشاد
مشرف / محمد أحمد الدسوقى
مناقش / إبراهيم محمود الحناوي
مناقش / مجدي زكريا رشاد
مناقش / سمير الدسوقي الموجي
الموضوع
Electronic data processing - Distributed processing. Computer networks.
تاريخ النشر
2018.
عدد الصفحات
91 p. :
اللغة
الإنجليزية
الدرجة
ماجستير
التخصص
Computer Networks and Communications
تاريخ الإجازة
1/7/2019
مكان الإجازة
جامعة المنصورة - كلية الحاسبات والمعلومات - علوم الحاسب
الفهرس
Only 14 pages are availabe for public view

Abstract

Modeling a distributed system is a difficult task, facing many challenges. Firstly, some applications have special requirements for high communication. According to the interaction model, there are many various problems related to the communication performance such as workload, and network delay.
Concerning the failure model, there could be many internal problems such as non-synchronized clocks.
So, it is an urgent need to devise new topology for efficient network communication. On the other hand, it is necessary to determine the minimal set of algorithms to cover the functionality of a distributed system.
This thesis proposes a new topology called the ”FIRM” based on analysis of complex systems.
Mathematical and visual analysis of this topology shows its efficiency in terms of network communication.
For n nodes, the mesh requires n (n-1)/2 of links, the ring requires n of links, the star requires n-1 of links, the bus requires (2n-1) of links, and the firm requires (n-1) of links.
Comparing the topologies, in terms of hops at N=100 nodes, shows the following:
• The Mesh requires 1 hop but is not feasible to build full connection because of high cost of cabling.
• The Ring requires 50 hops in average , but it will be slow.
• The Star requires 2 hopes, but core has bottle neck and represents a single point of failure.
• The Bus requires 50 hops in average , but not suitable in large networks because if cable fails then, whole network fails (no fault degradation), Slow and Limited cable length
• The Firm (proposed topology) requires exactly 4 hops (at maximum) but in regular 2 hops, high speed, no bottle neck and has fault degradation.
Then, the thesis provides a detailed scrutiny of the realization of this topology in hardware and network specification using packet tracer.