信息中心自组网中基于内容动态热度的缓存策略

Content Dynamic Popularity-Based Cache Strategy in Information Centric Ad Hoc Networks

摘要:

针对信息中心自组网的内容高效分发问题，提出了一种基于内容动态热度的缓存策略；阐述了信息中心自组网的特点与缓存策略的核心问题，并以降低网内负载、传输时延以及提高缓存命中率为目标构建系统架构；提出了内容动态热度的计算方式以及热点值的定义，改进了兴趣包与数据包的数据结构，并设计了相应的处理流程；仿真结果表明，基于内容动态热度的缓存策略，在高缓存占比的情况下，缓存命中率可以稳定达到87%；与传统的随机缓存与基于静态优先级的缓存策略相比，缓存命中率分别可以提升约12%和7%，全网平均负载相比随机缓存策略降低约40%；在高负载环境下，在保持低内容获取时延的同时，端到端时延增速显著降低；整体提升了信息中心自组网的内容分发效率与服务质量。

Abstract:

To address the challenge of efficient content distribution in information-centric ad hoc networks (ICANs), a dynamic content caching strategy based on dynamic content popularity is proposed.The distinctive characteristics of ICANs and the core challenges inherent in caching strategies are elaborated. A system architecture is developed with three optimized key objectives: minimizing intra-network traffic load and transmission latency, while maximizing the cache hit ratio. A dynamic content popularity metric is introduced, a formal definition of the popularity value is provided, the data structures of Interest and Data packets are improved, and the corresponding processing workflows are designed accordingly.Simulation results demonstrate that the proposed dynamic popularity-based content caching strategy achieves a stable cache hit ratio of 87% under high cache proportions. Compared with traditional random caching and static priority-based caching strategies, the proposed strategy improves the cache hit ratio by approximately 12% and 7%, respectively. Meanwhile, it reduces the average network load by about 40% compared to the random caching strategy. Furthermore, in high-load environments, the strategy maintains low content retrieval latency while significantly mitigating the growth rate of end-to-end delay. Overall, the dynamic content caching strategy effectively enhances the content distribution efficiency and quality of service (QoS) of ICANs.