International Journal of Science, Engineering and TechnologyAuthors: Haritha Bhuvaneswari Illa
Abstract: The evolution of virtualized cloud infrastructures has introduced significant challenges in maintaining efficient and reliable network routing performance across dynamically scalable, multi-tenant environments. This study presents a detailed comparative analysis of four well established routing protocols Ad hoc On Demand Distance Vector (AODV), Dynamic Source Routing (DSR), Optimized Link State Routing (OLSR), and Destination Sequenced Distance Vector (DSDV) within a simulated virtualized cloud framework. The experimental design integrates CloudSim for modelling cloud resource behaviour and Network Simulator 2 (NS2) for network level routing analysis, enabling concurrent evaluation of both virtualization and communication dynamics. The research investigates how virtualization overhead, virtual machine (VM) density, and workload intensity (light, moderate, and heavy) influence routing performance metrics, including throughput, end to end (E2E) delay, packet delivery ratio (PDR), routing overhead, and jitter. Simulation results reveal that virtualization introduces measurable delays and bandwidth contention due to shared I/O, hypervisor scheduling, and virtual switch buffering. Among the evaluated protocols, AODV consistently exhibited superior adaptability, achieving the highest average throughput (≈51 Mbps) and PDR (≈79%) with minimal delay and jitter across all scenarios. OLSR maintained stable delivery and predictable performance but incurred the highest routing overhead due to continuous topology updates. DSR demonstrated the lowest control traffic yet suffered from cache staleness under heavy load, while DSDV showed slower convergence and reduced efficiency in dynamic topologies. The findings establish that reactive routing protocols such as AODV and DSR outperform proactive routing protocols like OLSR and DSDV in virtualized cloud systems, primarily due to their ability to adapt dynamically to topological changes and VM migrations. This study identifies AODV as the most suitable protocol for large scale virtualized clouds requiring high adaptability and reliability. Furthermore, the results highlight the potential of integrating Software Defined Networking (SDN) and Network Function Virtualization (NFV) to enhance dynamic control, scalability, and Quality of Service (QoS) in future cloud routing frameworks
DOI: http://doi.org/10.5281/zenodo.17528825
