Content Popularity Prediction via Federated Learning in Cache-Enabled Wireless Networks

With the rapid development of networks,users are increasingly seeking richer and high-quality content experience,and there is an urgent need to develop efficient content caching strategies to improve the content distribution efficiency of caching.Therefore,it will be an effective solution to combine content popularity prediction based on machine learning(ML)and content caching to enable the network to predict and analyze popular content.However,the data sets which contain users’private data cause the risk of privacy leakage.In this paper,to address this challenge,we propose a privacy-preserving algorithm based on federated learning(FL)and long short-term memory(LSTM),which is referred to as FL-LSTM,to predict content popularity.Simulation results demonstrate that the performance of the proposed algorithm is close to the centralized LSTM and better than other benchmark algorithms in terms of privacy protection.Meanwhile,the caching policy in this paper raises about 14.3%of the content hit rate.

Analysis and Prediction of Content Popularity for Online Video Service:A Youku Case Study

Understanding the characteristics and predicting the popularity of the newly published online videos can provide direct implications in various contexts such as service design, advertisement planning, network management and etc. In this paper, we collect a real-world large-scale dataset from a leading online video service provider in China, namely Youku. We first analyze the dynamics of content publication and content popularity for the online video service. Then, we propose a rich set of features and exploit various effective classification methods to estimate the future popularity level of an individual video in various scenarios. We show that the future popularity level of a video can be predicted even before the video's release, and by introducing the historical popularity information the prediction performance can be improved dramatically. In addition, we investigate the importance of each feature group and each feature in the popularity prediction, and further reveal the factors that may impact the video popularity. We also discuss how the early monitoring period influences the popularity level prediction. Our work provides an insight into the popularity of the newly published online videos, and demonstrates promising practical applications for content publishers,service providers, online advisers and network operators.

A Cache Replacement Policy Based on Multi-Factors for Named Data Networking

Named Data Networking(NDN)is one of the most excellent future Internet architectures and every router in NDN has the capacity of caching contents passing by.It greatly reduces network traffic and improves the speed of content distribution and retrieval.In order to make full use of the limited caching space in routers,it is an urgent challenge to make an efficient cache replacement policy.However,the existing cache replacement policies only consider very few factors that affect the cache performance.In this paper,we present a cache replacement policy based on multi-factors for NDN(CRPM),in which the content with the least cache value is evicted from the caching space.CRPM fully analyzes multi-factors that affect the caching performance,puts forward the corresponding calculation methods,and utilize the multi-factors to measure the cache value of contents.Furthermore,a new cache value function is constructed,which makes the content with high value be stored in the router as long as possible,so as to ensure the efficient use of cache resources.The simulation results show that CPRM can effectively improve cache hit ratio,enhance cache resource utilization,reduce energy consumption and decrease hit distance of content acquisition.

Analyzing the dynamics of online video popularity

Given the large volume of video content and the diversity of user attention, it is of great importance to understand the characteristics of online video popularity for technological, economic and social reasons. In this paper, based on the data collected from a leading online video service provider in China, namely Youku, the dynamics of online video popularity are analyzed in-depth from four key aspects： overall popularity distribution, individual popularity distribution, popularity evolution pattern and early-future popularity relationship. How the popularity of a set of newly upload videos distributes throughout the observation period is first studied. Then the notion popularity distributions of individual videos are carefully studied. of active days is proposed, and the per-day and per-hour Next, how the popularity of an individual video evolves over time is investigated. The evolution patterns are further defined according to the number and temporal locations of popularity bursts, in order to describe the popularity growth trend. At last, the linear relationship between early video popularity and future video popularity are examined on a log-log scale. The relationship is found to be largely impacted by the popularity evolution patterns. Therefore, the specialized models are proposed to describe the correlation according to the popularity evolution patterns. Experiment results show that specialized models can better fit the correlation than a general model. Above all, the analysis results in our work can provide direct help in practical for the interested parties of online video service such as service providers, online advisers, and network operators.

Joint model of user check-in activities for point-of-interest recommendation

With the rapid development of location-based networks, point-of-interest（POI） recommendation has become an important means to help people discover interesting and attractive locations, especially when users travel out of town. However, because users only check-in interaction is highly sparse, which creates a big challenge for POI recommendation. To tackle this challenge, we propose a joint probabilistic generative model called geographical temporal social content popularity（GTSCP） to imitate user check-in activities in a process of decision making, which effectively integrates the geographical influence, temporal effect, social correlation, content information and popularity impact factors to overcome the data sparsity, especially for out-of-town users. Our proposed the GTSCP supports two recommendation scenarios in a joint model, i.e., home-town recommendation and out-of-town recommendation. Experimental results show that GTSCP achieves significantly superior recommendation quality compared to other state-of-the-art POI recommendation techniques.