Self-Sizing Techniques for Locally Controlled Networks

Self-Sizing;Global;Local

Nalatwad, Srikant S ; Michael Devetsikiotis, Committee Chair,Harry Perros, Committee Member,Mihail L Sichitiu, Committee Member,Khaled A. Harfoush, Committee Member,Nalatwad, Srikant S ; Michael Devetsikiotis ; Committee Chair ; Harry Perros ; Committee Member ; Mihail L Sichitiu ; Committee Member ; Khaled A. Harfoush ; Committee Member

University:North Carolina State University

关键词: Self-Sizing;    Global;    Local;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/4380/etd.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

The Internet was designed to provide best effort service for delivery of data packets and to run virtually across any network transmission and system platform. Its exponential growthhas turned it into a multiservice complex network of heterogeneous elements with dynamically changing traffic conditions. To regulate such a large scale network it is necessary to place intelligence in the nodes. Network control should be decentralized to make such a system reliable and manageable. It is necessary to find simple local rules and strategies that can produce purposeful and coherent behavior. These control mechanisms must be adaptive to effectively respond to continually varying network conditions. Such adaptive, distributed, localized mechanisms would provide a scalable solution for controlling these large networks. Our comprehensive study on QoS developments in the Internet, reveals the necessity and requirement of a new QoS framework which provides absolute guarantees to the underlying traffic. We propose an innovative self-sizing framework for locally controlled networks such as the Internet,which can support the stringent requirements of interactive applications. A "self-sizing" network can allocate link/switch capacity automatically and adaptively using online traffic data. Our unified, critical and comparative analysis of online resource allocation algorithms of two different classical approaches, leads us to a novel adaptive wavelet predictor. Our results show that by performing online resource allocation at each node based on their local knowledge, we can achieve considerable bandwidth savings and also satisfy QoS at the packet level. We further discover that by making some of the nodes aware of their neighbors resource availability, higher self-sizing gains can be attained.