Context-Aware Proactive Content Caching With Service Differentiation in Wireless Networks

Abstract

Content caching in small base stations or wireless infostations is considered to be a suitable approach to improve the efficiency in wireless content delivery. Placing the optimal content into local caches is crucial due to storage limitations, but it requires knowledge about the content popularity distribution, which is often not available in advance. Moreover, local content popularity is subject to fluctuations since mobile users with different interests connect to the caching entity over time. Which content a user prefers may depend on the user's context. In this paper, we propose a novel algorithm for contextaware proactive caching. The algorithm learns context-specific content popularity online by regularly observing context information of connected users, updating the cache content and observing cache hits subsequently. We derive a sublinear regret bound, which characterizes the learning speed and proves that our algorithm converges to the optimal cache content placement strategy in terms of maximizing the number of cache hits. Furthermore, our algorithm supports service differentiation by allowing operators of caching entities to prioritize customer groups. Our numerical results confirm that our algorithm outperforms state-of-the-art algorithms in a real world data set, with an increase in the number of cache hits of at least 14%.

A Novel Caching Policy with Content Popularity Prediction and User Preference Learning in Fog-RAN

“…From the MovieLens, we select the accessing dataset of the chosen 30 users requested from January 01, 2010 to October 17, 2016, where the first part of the accessing dateset, whose requesting date is from January 01, 2010 to December 31, 2015, is used for initializing the user preference, while the second part of the accessing dateset, whose requesting date is from January 01, 2016 to October 17, 2016, is used for the performance evaluation. Considering that users will generally comment on the movie after they have watched it, we take the movie rating from a user as the request for this movie [9], [10]. In our simulations, we set the number of the considered F-APs to 3, the finite time horizon to 6984 hours, the monitoring cycle to 1 hour, and the predefined threshold to 0.2, respectively.…”

“…Meanwhile, the transfer learning approach has a poor performance for the case of low information correspondence ratio. In [9], the cache content placement problem was modeled as a contextual multi-arm bandit problem and an online policy was presented to learn the content popularity. This policy learns the content popularity independently across contents whereas ignores the content similarity and impact of user preference on content popularity, thereby resulting in high training complexity and slow learning speed.…”

Dynamic service deployment for budget‐constrained mobile edge computing

“…Prabh and Abdelzaher developed an application service to search the optimal locations for caching data and reducing packet transmission power. Müller et al put forward a new approach to learn the knowledge regarding context‐specific popularity, and then updated content cached on each BS adaptively. Different from data/content caching where the major concern is concentrated on the caching under server's storage limit, a feasible service deployment should take into account not only storage capacity but also computing power.…”

“…Prabh and Abdelzaher 21 developed an application service to search the optimal locations for caching data and reducing packet transmission power. Müller et al 22 put forward a new approach to learn the knowledge regarding context-specific popularity, and then updated content cached on each BS adaptively.…”

Optimal Pricing of Internet of Things: A Machine Learning Approach

“…Z. Han is with the University of Houston, TX, USA 77004 (email: zhan2@uh.edu), and also with the Department of Computer Science and Engineering, Kyung Hee University, South Korea 446-701. 1 For the rest of this paper, we use "machine learning-based IoT services" and "services" interchangeably. required to ensure maximum profits of firms by achieving optimal utilization of the IoT resources and optimal subscription fees for services.…”