Cocktail Edge Caching: Ride Dynamic Trends of Content Popularity with Ensemble Learning

被引:15
作者
Zong, Tongyu [1 ]
Li, Chen [1 ]
Lei, Yuanyuan [1 ]
Li, Guangyu [1 ]
Cao, Houwei [2 ]
Liu, Yong [1 ]
机构
[1] NYU, Tandon Sch Engn, New York, NY 10003 USA
[2] New York Inst Technol, Dept Comp Sci, Old Westbury, NY 11568 USA
来源
IEEE CONFERENCE ON COMPUTER COMMUNICATIONS (IEEE INFOCOM 2021) | 2021年
关键词
edge caching; video; deep reinforcement learning; LSTM; REINFORCEMENT; PREDICTION;
D O I
10.1109/INFOCOM42981.2021.9488910
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Edge caching will play a critical role in facilitating the emerging content-rich applications. However, it faces many new challenges, in particular, the highly dynamic content popularity and the heterogeneous caching configurations. In this paper, we propose Cocktail Edge Caching, that tackles the dynamic popularity and heterogeneity through ensemble learning. Instead of trying to find a single dominating caching policy for all the caching scenarios, we employ an ensemble of constituent caching policies and adaptively select the best-performing policy to control the cache. Towards this goal, we first show through formal analysis and experiments that different variations of the LFU and LRU polices have complementary performance in different caching scenarios. We further develop a novel caching algorithm that enhances LFU/LRU with deep recurrent neural network (LSTM) based time-series analysis. Finally, we develop a deep reinforcement learning agent that adaptively combines base caching policies according to their virtual hit ratios on parallel virtual caches. Through extensive experiments driven by real content requests from twod large video streaming platforms, we demonstrate that CEC not only consistently outperforms all single policies, but also improves the robustness of them. CEC can be well generalized to different caching scenarios with low computation overheads for deployment.
引用
收藏
页数:10
相关论文
共 45 条
[1]  
[Anonymous], 2018, Cisco Visual Networking Index: Forecast and Trends, 2017-2022
[2]  
[Anonymous], 1998, IMPROVING WWW PROXIE
[3]   A STUDY OF REPLACEMENT ALGORITHMS FOR A VIRTUAL-STORAGE COMPUTER [J].
BELADY, LA .
IBM SYSTEMS JOURNAL, 1966, 5 (02) :78-&
[4]  
Bell R., 2008, AT T RES
[5]  
Brown R., 2012, Introduction to Random Signals and Applied Kalman Filtering, V4th
[6]   A FIFO-based buffer management approach for the ATM GFR services [J].
Chan, CT ;
Hu, SC ;
Wang, PC ;
Chen, YC .
IEEE COMMUNICATIONS LETTERS, 2000, 4 (06) :205-207
[7]  
Coffman E. G., 1973, Operating Systems Theory
[8]  
DAN A, 1990, PERF E R SI, V18, P143, DOI 10.1145/98460.98525
[9]  
Fan Q., 2020, PA CACHE LEARNING BA
[10]  
Fofack N. C., 2012, 2012 6th International ICST Conference on Performance Evaluation Methodologies and Tools (VALUETOOLS 2012). Proceedings, P1, DOI 10.4108/icst.valuetools.2012.250250