Placing selected content at the edge of the network close to the users, known as caching, is an important technique to improve the efficiency of content delivery in wireless networks. In this paper, we consider caching in a cloud radio access network (C-RAN) in which the primary fronthaul link operates in the mmWave range and may switch to microwave frequencies in the case of blockage. We aim to minimize the average long-term network cost by optimizing dynamic fetching and caching decisions. Importantly, we consider the realistic case of user request distributions and blockage rates being a priori unknown and not necessarily stationary. We introduce change point detection (CPD) to detect significant changes in the environment; we couple this step with reinforcement learning (RL): our key contribution, the proposed change point detection assisted reinforcement learning (CPRL) algorithm learns the environment and (re-)optimizes the caching policy to solve the associated Markov decision process (MDP) problem. Essentially, CPD allows our learning algorithm to adapt its caching strategy to the new environment which shows faster convergence. The numerical results show that our proposed approach improves the efficiency of caching in wireless networks, making it more adaptable to changing request patterns over time.

Placing selected content at the edge of the network close to the users, known as caching, can lower network latency and congestion in the fronthaul link. Unlike most works that assume a fixed or limited variation in file popularities, to better address user requests, we consider a time-varying popularity resulting in hidden-mode Markov decision processes. In fact, each mode captures environmental changes, and we optimize the fetching and dropping (of files) decisions to minimize a long-term network cost in a cloud radio access network. Importantly, the primary fronthaul link is a millimeter (mmWave) link with large capacity supported by a microwave backup link in case of blockage. Since caching decisions are coupled over time and can affect the future, we introduce a dynamic programming approach to solve the caching problem. We approximate the future cost of each cache state in each mode. To reduce the complexity of calculating the future cost, we introduce two approximation approaches and illustrate the accuracy of the approximations under different environmental conditions. Finally, our simulation results confirm the effectiveness of our proposed algorithm in finding effective caching and fetching decisions to lower the total network cost while dealing with time-varying popularities.

Abstract In this paper, the problem of opportunistic spectrum sharing for the next generation of wireless systems empowered by the cloud radio access network (C‐RAN) is studied. More precisely, low‐priority users employ cooperative spectrum sensing to detect a vacant portion of the spectrum that is not currently used by high‐priority users. The authors' aim is to maximize the overall throughput of the low‐priority users while guaranteeing the quality of service of the high‐priority users. This objective is attained by optimally adjusting spectrum sensing time, with respect to target probabilities of detection and false alarm, as well as dynamically allocating C‐RAN resources, that is, powers, sub‐carriers, remote radio heads, and base‐band units. To solve this problem, which is non‐convex and NP‐hard, a low‐complex iterative solution is proposed. Numerical results demonstrate the necessity of sensing time adjustment as well as effectiveness of the proposed solution.

This paper addresses the problem of sub-carrier and user association in a downlink cloud based radio access network (C- RAN), considering fronthaul and access radio frequency (RF) links and orthogonal frequency division multiple access (OFDMA). Our problem is most relevant to scenarios where bandwidth resources must be shared between the fronthaul and access links. In order to assign users to their appropriate cells and to allocate frequency resources, we maximize the sum throughput. Importantly, we consider fronthaul and inter-cell interference. The resulting optimization problem is based on joint fronthaul and access frequency resource allocation and user association and is non-convex. To tackle the non-convexity of the problem, a successive convex approximation method is proposed. In order to guarantee integer solutions, we introduce a term, called virtual interference, into the problem formulation. Numerical results validate the effectiveness of proposed algorithm in jointly allocating resources of fronthaul and access links. The results confirm improved total network throughput by considering full interference scheme and sharing resources between fronthaul and access links.