Load Balancing Strategy for Large Scale Software Defined Networks

Title

Load Balancing Strategy for Large Scale Software Defined Networks

Creator

Kulkarni, Manasa

Contributor

Paulose, Joy and Goswami, Bhargavi

Description

Programmability has left its mark on every facet of business, with technology playing newlinean integral role. Social networking industry trends underscore technology s ubiquity in newlinenearly every business transaction. Traditional networks grapple with numerous challenges, rendering them ill-equipped to process and handle the demands of the modern newlinelandscape effectively. The lack of programming in these networks leads to stagnation, newlineinhibiting their ability to evolve or enhance performance. The advent of Software Defined Networks (SDN) has introduced increased flexibility into conventional networks, newlineopening avenues for creating innovative services. newlineSDN technology addresses challenges in large-scale networks, offering solutions for newlinehigh throughput, virtualization, fault detection, and load balancing, providing effective network management. The rapid expansion of network services and applications newlinein SDN environments demands sophisticated load-balancing solutions that adapt to newlinedynamic traffic patterns and varying service requirements. This study presents a pioneering algorithm, the Dynamic Load Balancing Algorithm (DLBA), which utilizes the newlineProgramming Protocol-independent Packet Processors (P4) language. The algorithm is newlinespecifically crafted to tackle the issues associated with optimizing traffic distribution in newlinethe data plane of SDN. newlineP4 programming language, recognized as one of the most robust languages, addresses newlinethe limitations of traditional networking, enhancing programmability and agility by newlinedistributing the load across the network. The research implements a novel quotDynamic newlineLoad Balancing Algorithmquot using the P4 language to instill dynamism and achieve load newlinebalance in large-scale networks. The P4-based implementation showcases dynamicity, scalability, flexibility, and adaptability. This research commences with thoroughly newlineexamining existing load-balancing algorithms implemented using the P4 language, followed by a comparative analysis between these algorithms and DLBA.

Source

Author's Submission

Date

2024-01-01

Publisher

Christ(Deemed to be University)

Subject

Computer Science

Rights

Open Access

Relation

61000345

Format

PDF

Language

English

Type

PhD

Identifier

http://hdl.handle.net/10603/577509