Caching with Unknown Popularity Profiles in Small Cell Networks

Abstract

Abstract-A heterogenous network is considered where the base stations (BSs), small base stations (SBSs) and users are distributed according to independent Poisson point processes (PPPs). We let the SBS nodes to posses high storage capacity and are assumed to form a distributed caching network. Popular data files are stored in the local cache of SBS, so that users can download the desired files from one of the SBS in the vicinity subject to availability. The offloading-loss is captured via a cost function that depends on a random caching strategy proposed in this paper. The cost function depends on the popularity profile, which is, in general, unknown. In this work, the popularity profile is estimated at the BS using the available instantaneous demands from the users in a time interval [0, τ ]. This is then used to find an estimate of the cost function from which the optimal random caching strategy is devised. The main results of this work are the following: First it is shown that the waiting time τ to achieve an ǫ > 0 difference between the achieved and optimal costs is finite, provided the user density is greater than a predefined threshold. In this case, τ is shown to scale as N 2 , where N is the support of the popularity profile. Secondly, a transfer learning-based approach is proposed to obtain an estimate of the popularity profile used to compute the empirical cost function. A condition is derived under which the proposed transfer learning-based approach performs better than the random caching strategy.

Transfer Learning for Future Wireless Networks: A Comprehensive Survey

“…2) Content Popularity Prediction at Macro BSs: The use of inductive TL to predict the content popularity in heterogeneous networks considering the minimum training time is proposed in [163], [164]. Instead of estimating the content popularity at each SBS, the SBS only incorporates a random caching strategy.…”

“…Although the works in [161], [162] demonstrate the superiority of using TL approaches, they assume that the content popularity is predicted independently at each SBS, leading to the redundant caching problem. Furthermore, the authors in [163]- [165] consider that the content popularity estimation is performed at the centralized BS, thereby increasing the workload and operational cost at the BS. To address these challenges, the authors in [166] propose an inductive TLbased proactive caching mechanism under the cooperative caching scenario among edge servers as shown in Fig.…”

Cache Placement Optimization in Mobile Edge Computing Networks with Unaware Environment -- An Extended Multi-armed Bandit Approach

“…For example, [3,4] employ the deep learning-based prediction by firstly collecting users' requests as the training data. In [5,6], transfer learning is introduced to estimate the popularity, which is further used to design the cache placement strategy. Specifically, [6] studies the cache placement in a heterogeneous network where the content popularity information is unaware.…”

“…In [5,6], transfer learning is introduced to estimate the popularity, which is further used to design the cache placement strategy. Specifically, [6] studies the cache placement in a heterogeneous network where the content popularity information is unaware. [5] discusses content correlation and information transition between periods and uses the auto-regressive (AR) model to predict the users' requests.…”

The role of caching in next generation cellular networks: A survey and research outlook

“…Such networks offer an important applications area for local content caching by using the SBSs as cache storage units at the edge of the network. In other works, small cell network proactive caching has been studied. A distributed cache placement problem is recently studied in the work of Shanmugam et al with an aim to reduce the delay in delivering the files to the end user; whereas, in the work of Blasco and Gunduz, a single SBS cache content placement problem was addressed from a reinforcement learning perspective.…”