Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 46126 (19)
- University: Sharif University of Technology
- Department: Computer Engineering
- Advisor(s): Hemmatyar, Ali Mohammad Afshin
- Abstract:
- We call medium and large networksSDN, which they can managed by centralized software. In this networks instead ofmaking routing decisions byhardware, software is making decisions by using pre-defined conditions and with each packet behave differently. With the increasing development of real-time services such as voice services the importance of network monitoring system is increasing day after day. IN order to ensure quality of service, network service providers must measure network status with metrics such as delay and packet loss rate. Obtaining accurate and meaningful statistics from network status, provide abilities to service providers to measuretheir network health and estimate amount of servicedegradation in congestion time. Also they can use this information to create better route decisions.Solutions that have been proposed to measure the required standards is a collection of several different methods, each of which requires a separate infrastructure. Besides being costly and need to be installed. Furthermore, none of the proposed methods cannot meet all the requirements of a complete monitoring solution. Because they may not be accurate or causes overhead on the network and eliminate network stability.SDN networks can be used as an integrated solution for the measurement criteria that needed to be examined. The SDN networks are still in the early stages of their development and have not been much discussed.In this thesis we focused on tradeoffs between accuracy and network overhead for retrievestatistics, and we will represent a framework for software defined networking. The proposed way provide interface for collecting information in different aggregation levels. Alsowe will use an algorithm with the ability to adapt the network conditions and provide accurate information in real time without adding unreasonable overhead on the network.In this regard, we use floodlight's application programming interface, to implement the framework.After numerous tests performed, respectively, it’s clear that the proposed method for collecting statistics is more accurate than other algorithms, and yet, the overhead produced by the method in contrast of direct measurement techniques, is about 50% lower
- Keywords:
- Routing ; Simulators ; Efficiency ; Software Defined Networking (SDN) ; OpenFlow Protocol
Digital Object List
-
محتواي کتاب
- view
Bookmark