Dynamic Programming Based Content Placement Strategy for 5G and Beyond Cellular Networks

Abstract

Optimal caching strategies of popular contents in heterogeneous cellular networks are studied. The increasing demand for data traffic by users of the wireless network can be handled by rapaciously caching most frequently accessed contents by users. Hence, we propose an efficient popular content placement strategy, the first step in the content caching process, typically for popular video files. To do so, we introduce a novel approach for caching popular contents. This caching strategy follows a dynamic programming approach to tackle the optimization complexity of selecting most popular files among a wide range of files, under certain constraints. The proposed strategy gives the combination of popular files to be cached that maximizes the optimal cache hit probability with a pseudo-polynomial time complexity. To that end, we used the well-known resourcing algorithm, called the 0/1-Knapsack problem, assuming that files are cached without partitioning.

Performance evaluation for new web caching strategies combining LRU with score based object selection

“…The bound on meets the optimum value hit ratio, if L top scored objects exactly fit into the cache ( q = 0). Knapsack solutions for IRM request pattern are addressed manifold [4] [34][59] [82][92] [93][99] [116]. Tatarinov [116] suggests a knapsack algorithm with documents D in the cache being sorted due to "value(D) = benefit(D) / size(D)" or in our notation: SD = vD / sD.…”

“…While the analysis becomes more complex for flexible and advanced caching methods, bounds for optimum caching efficiency have been derived from Belady's bound to general min-cost flow and knapsack solutions, which fully cover scorebased strategies [4][7] [11][19] [58][59] [92][93] [112] [116]. Time-to-live (TTL) caching emerged as another branch of strategies for web applications, where each object in a cache is valid only for a limited time.…”

“…(1) General Bounds on the Cache Hit and Value Ratio We start from knapsack bounds for optimum static caches for IRM requests [4][92] [93][99] [116]. They are extended at first to varying request rates over time, and finally to arbitrary request pattern in 2-dimensional (2D-)knapsack solutions [19] [59].…”