China
Africa
Explore chapters and articles related to this topic
Cloud VR Service Platform Technologies
Published in Huaping Xiong, Dawei Li, Kun Huang, Mu Xu, Yin Huang, Lingling Xu, Jianfei Lai, Shengjun Qian, Cloud VR, 2020
Huaping Xiong, Dawei Li, Kun Huang, Mu Xu, Yin Huang, Lingling Xu, Jianfei Lai, Shengjun Qian
To enable quick download, items that are frequently requested can be cached on CDN nodes. These cached items can be dynamically adjusted depending on their popularity as well as available storage capacity. The most commonly used cache algorithms include Least Recently Used (LRU) and Least Frequently Used (LFU).[13] Meanwhile, its basic function is to determine whether to cache or delete an item by collecting and analyzing information such as access time, count, and frequency.
An approximate dynamic programming approach for collaborative caching
Published in Engineering Optimization, 2021
In the wireless caching systems, the cache optimization problem is hard to solve owing to the plethora of the requested contents, the dynamic nature of the requests and base stations having limited cache space. This problem becomes even more challenging because the vast majority of the generated data are video files, which have considerable size, and there are strict delivery deadlines. Thus, even when base stations determine the optimal caching policy independently, the cache optimization problem is still NP-hard as it can be mapped to a knapsack problem (Martello and Toth 1990). In practice, network operators can adopt simple cache update policies such as the Least Recently Used (LRU), Least Frequently Used (LFU), or other more advanced methods (Muller et al.2017; Bharath et al.2018; Abad et al.2019; Yang et al.2019) that are applied separately to each BS. These methods are intuitive and show good performance in cases of independent BS cache optimization, but do not work well in collaborative caching because the optimal replacement decisions also depend on the network topology.