Pricing, rate control and content placement for caching networks
Songqi Geng · 2023
This paper addresses the issue of pricing, rate control, and placement of content over a caching network. Consider a scenario in which both the network and its users are selfish. Users are subject to charges by network based on their request rate, quantified per unit of time and unit of flow. Specifically, the network seeks to maximize its revenue, while users aim to maximize their benefits and minimize their expenditures. Within this framework, an equilibrium point emerges between what can be termed as the "Network Problem" and the "Users Problem". At this equilibrium, both the network and the users reach their optimal solutions, and the solution leads to the maximization of the overall system utility. We say an equilibrium is global if the solution globally solves the "Network Problem" and the "User Problem". However, global equilibrium does not always exist, since the "Network Problem" is NP-hard. We provide necessary and sufficient conditions for the existence of global equilibrium. Nevertheless, we show that local equilibrium always exists, where the solution locally solves "Network Problem" and "User Problem". We also provide a method to find a local equilibrium with an optimality guarantee. Furthermore, we also discuss the rounding algorithm in such equilibrium of caching network.--Author's abstract