ذخيره‌سازي آگاه از محتوا در شبكه‌هاي مبتني بر نام داده

With advancements in technology an‎d the emergence of new applications, the conventional Internet architecture faces escalating challenges, highlighting the urgent necessity for a transition to the future architecture of the Internet an‎d the benefits of Information-Centric networks. These challenges stem from changes in the needs an‎d number of users, an‎d their usage patterns. Named Data Networks emerges as a leading an‎d reliable approach, offering a viable solution to meet growing user needs by focusing on the content itself rather than its physical location. A prominent characteristic of these networks is their ubiquitous in-network caching capability. This feature enables optimal content distribution across the network by decoupling content from its original source, consequently leading to a significant reduction in content access latency, link ban‎dwidth consumption, an‎d overall network traffic, thereby enhancing the Quality of Experience (QoE) for users. In the first phase of this research, a collaborative caching strategy based on the cooperation of adjacent nodes along the content delivery path was developed. This approach establishes virtual local clusters with distributed caching space an‎d minimal announcement message exchange. To optimally leverage the storage resources of these adjacent nodes along the delivery path, two novel modules, RIT an‎d NCT, are introduced. Subsequently, an adaptive caching threshold determination mechanism is proposed for each node, simultaneously considering content parameters, network topology, an‎d the dynamic state of the node. Finally, a decision mechanism regarding content storage at each node along the path with an adaptive approach based on the determined threshold is presented. This adaptive mechanism enhances storage efficiency across the network domain by preventing unnecessary content redundancy. In the second phase, aimed at overcoming the storage limitations of routers, a content-priority-based cache replacement policy is proposed. This policy assigns an Efficiency Factor (EF) to each content item stored in the cache, proportional to its influential parameters. A Hierarchical Analysis Process (AHP) is employed for optimal parameter weighting. This framework facilitates effective cache management an‎d precise selec‎tion of content to be evicted. Extensive simulation results demonstrate the superiority of the proposed approach over reference strategies (LCE, LCD, MCD, CL4M, an‎d ProbCache). In the best-case scenario, an improvement of 16.1% in the Cache Hit Ratio, a 27.6% reduction in average data delivery latency, an‎d a 4% improvement in path stretch were achieved. These accomplishments not only validate the high efficiency an‎d potential of the proposed approach in optimizing network performance but also represent a significant step forward in the development of future Content-Centric Networks.

6

157254