A new coordinated caching placement mechanism for VoD in named data networking
Despite significant advancements in Named Data Networking (NDN), the efficient handling of Video on Demand (VoD) traffic remains a challenge due to suboptimal caching strategies, resulting in increased End-to-End Delay (E2EDelay) and elevated server load. Existing mechanisms often fail to manage sim...
| Auteur principal: | |
|---|---|
| Format: | Thèse |
| Langue: | anglais anglais anglais |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | https://etd.uum.edu.my/12051/1/depositpermission.pdf https://etd.uum.edu.my/12051/2/s901914_01.pdf https://etd.uum.edu.my/12051/3/s901914_02.pdf https://etd.uum.edu.my/12051/ |
| Abstract | Abstract here |
| Résumé: | Despite significant advancements in Named Data Networking (NDN), the efficient handling of Video on Demand (VoD) traffic remains a challenge due to suboptimal caching strategies, resulting in increased End-to-End Delay (E2EDelay) and elevated server load. Existing mechanisms often fail to manage simultaneous online video requests effectively, particularly under dynamic network conditions. This study addresses these inefficiencies by proposing a novel coordinated caching placement mechanism tailored for VoD services in NDN environments. The first objective of the research is to evaluate the functional characteristics of seven existing caching placement mechanisms through simulation, identifying the features essential for optimal VoD content delivery. Building on these insights, a new mechanism, Coordination Caching Placement by Video Partitioning (CPVP), is introduced. CPVP aims to enhance caching efficiency, reduce latency, and improve user experience. CPVP utilises hypergraph theory to model cooperative caching behaviour across the network. It dynamically partitions video content and strategically stores segments along the path to clients using three neighbouring routers. The mechanism incorporates three integrated schemes: the Video Aggregation Scheme (VAS), which identifies the full video and determines cacheable routers; the Video Cache Decision Scheme (VCDS), which minimises redundancy and latency; and the Video Partitioning Scheme (VPS), which divides the video into three blocks for distributed caching. Simulations are conducted using ndnSIM 2.0 across four network topologies, evaluating four key performance metrics. Results demonstrate that CPVP outperforms existing mechanisms such as Leave Copy Everywhere (LCE), Leave Copy Down (LCD), and probabilistic caching (Prob). Specifically, CPVP achieves a 22.3% reduction in average E2EDelay and a 56.5% improvement in cache hit ratio on the GEANT topology. This research contributes to the development of efficient caching strategies in NDN and offers practical improvements in VoD streaming quality and network performance |
|---|