Virtual network function allocation based on practical resource characterization = 실망환경 자원 특성을 기반으로 한 가상망 할당 연구

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 34
  • Download : 0
The major role of a network provider is to ensure quality of service (QoS) in the network. In order to do so, acquiring and testing new physical devices are routinely required. Provisioning new physical devices typically introduces high capital expenditure (CAPEX) and operational expenditure (OPEX) for the network provider. Therefore, finding a cost-effective method that can reduce the cost of testing and implementation cycle is critically required. In this regard, network function virtualization (NFV) can be one of the most suitable approaches to solve those problems. The main idea of NFV is decoupling the network functions from the physical equipment so that the functions can be installed into the physical commodity devices. The purpose of NFV is to reduce the deployment cost of a large network. NFV also gives flexibility to the network provider to do the service provisioning. Compared to NFV, Software-defined networking decouples data plane and control plane. SDN uses the centralized controller to manage all information that has been exchanged in the data plane. SDN gives open interfaces and brings unified programmable control that can be used to automatically configure the network. One of the main challenges in the NFV is resource allocation. Resource allocation is the method that explains how to compose and allocate network function into the physical substrate network efficiently. The problem is assigning virtual resources to substrate physical network. There are many types of NFVs. In this thesis, extensive experimental studies were conducted to characterize the CPU resources of firewall, proxy, web application firewall, and intrusion detection system (IDS) as a function of input load and configuration rules. Moreover, we use iptables as a firewall and NAT gateway, Squid as a proxy, ModSecurity as a web application firewall, and Snort as an IDS. From the experiment, we modeled CPU resource utilization as a function of input load and VNF configuration rules, which was used in the proposed heuristic algorithm to solve the resource allocation problem. We also compared the performance of the proposed heuristic algorithm with the ILP-based algorithm using an ILP-solver.
Advisors
Rhee, June Kooresearcher이준구researcher
Description
한국과학기술원 :전기및전자공학부,
Publisher
한국과학기술원
Issue Date
2018
Identifier
325007
Language
eng
Description

학위논문(석사) - 한국과학기술원 : 전기및전자공학부, 2018.2,[iii, 33 p. :]

Keywords

Network Function Virtualization▼aSoftware-defined Networking▼aresource characterization▼aresource allocation▼ainteger linear programming▼aheuristic algorithm; 네트워크 기능 가상화▼a소프트웨어 정의 네트워크▼a자원 정의▼a자원 할당▼a정수 선형 계획법▼a휴리스틱 알고리즘

URI
http://hdl.handle.net/10203/266905
Link
http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=733977&flag=dissertation
Appears in Collection
EE-Theses_Master(석사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0