Reducing Content Retrieval Latency in ICN
Fariborz Derakhshan, Andreas Timm‐Giel, Ramón Agüero · 2025
To minimize content access latency within Information-centric Networking (ICN), its default caching strategy, “Leave Copy Everywhere” (LCE), stores contents at each node along their delivery trajectory. However, this method inadvertently degrades caching efficiency due to the application of content replacement policies such as Least Recently Used (LRU). An indiscriminate replication of contents leads to lower cache hit ratios, decreased bandwidth utilization, and increased strain on content servers, resulting in longer content retrieval times. Thus, a nuanced modulation of contents' caching within the network is imperative to extend their proximity to their respective clientele. Identifying and strategically replicating high-demand contents across an optimum set of network nodes can maximize cache utilization and minimize cache pollution, notably improving network content accessibility and reducing retrieval times. We propose a caching strategy that dynamically, deterministically, and organically replicates the most popular contents to the optimum central cache nodes in the network, thus shortening the average path length that both requests and contents must traverse. Our caching strategy significantly outperforms LCE, achieving more than a 6 % reduction in content retrieval path lengths. This paper comprehensively evaluates the proposed algorithm, elucidating its functionality and consequential repercussions on content redundancy, hit rate, and access time, which are paramount in the context of 6G systems. Leveraging the core principles of the introduced caching strategy, such as intelligent content distribution and proximity-based data access, can be vital for realizing Ultra-Reliable Low Latency Communications (URLLC) and enhanced Mobile Broadband (eMBB).