Enabling In-Network Caching in Traditional IP Networks: Selective Router Upgrades and Cooperative Cache Strategies

Jiangping Han,Kaiping Xue,Jian Li,Jing Zhang,Zixuan Huang,David S. L. Wei
DOI: https://doi.org/10.1109/tnse.2023.3305195
IF: 6.6
2023-01-01
IEEE Transactions on Network Science and Engineering
Abstract:Enabling in-network caching in a traditional IP network by progressively adding cache-enabled nodes into the network can provide a variety of advantages, such as efficient content distribution and improved network resource utilization. However, to maximize the performance, two crucial issues have to be addressed: 1) due to a limited budget, only selective nodes could be upgraded, and thus which nodes should be chosen is of importance and 2) how to decide the corresponding cache strategy in an upgraded hybrid network is also important. In this paper, we first formulate a problem of Selective Router Upgrade (SRU) aiming at selecting nodes to be upgraded such that both average access delay and hit ratio are optimized for the best experience. We prove that SRU is an NP-hard problem, and then propose a ( $1-1/e$ )-approximation algorithm for it. Based on the model of SRU, we further propose Local Replacement (LR) and Neighbor Cooperative Caching (NCC) strategies to provide caching service in an upgraded hybrid network. Specifically, LR considers the cache behavior in a local router for both low latency and high hit ratio. NCC allows routers to obtain contents from their neighbor nodes and further decides the caching replacement policy based on neighbor nodes' caching capabilities. In a word, LR and NCC help expand the overall caching space and improve the content delivery efficiency by exploring the geometrical vicinity of nodes. Extensive simulation results show that our schemes can significantly improve the network in terms of access delay and hit ratio.
What problem does this paper attempt to address?