Rajarshi Bhattacharjee
Abhishek Sinha
Skip Abstract Section
Abstract
Optimal caching of files in a content distribution network (CDN) is a problem of fundamental and growing commercial interest. Although many different caching algorithms are in use today, the fundamental performance limits of network caching algorithms from an online learning point-of-view remain poorly understood to date. In this paper, we resolve this question in the following two settings: (1) a single user connected to a single cache, and (2) a set of users and a set of caches interconnected through a bipartite network. Recently, an online gradient-based coded caching policy was shown to enjoy sub-linear regret. However, due to the lack of known regret lower bounds, the question of the optimality of the proposed policy was left open. In this paper, we settle this question by deriving tight non-asymptotic regret lower bounds in both of the above settings. In addition to that, we propose a new Follow-the-Perturbed-Leader-based uncoded caching policy with near-optimal regret. Technically, the lower-bounds are obtained by relating the online caching problem to the classic probabilistic paradigm of balls-into-bins. Our proofs make extensive use of a new result on the expected load in the most populated half of the bins, which might also be of independent interest. We evaluate the performance of the caching policies by experimenting with the popular MovieLens dataset and conclude the paper with design recommendations and a list of open problems.
- [n.d.]. MovieLens 25M Dataset. https://grouplens.org/datasets/movielens/. Accessed: 01--19--2019.Google Scholar
- Jacob Abernethy, Chansoo Lee, Abhinav Sinha, and Ambuj Tewari. 2014. Online linear optimization via smoothing. In Conference on Learning Theory. 807--823.Google Scholar
- Jacob Abernethy and Alexander Rakhlin. 2008. Optimal strategies and minimax lower bounds for online convex games. In In Proceedings of the Nineteenth Annual Conference on Computational Learning Theory.Google Scholar
- Charu Aggarwal, Joel L Wolf, and Philip S. Yu. 1999. Caching on the world wide web. IEEE Transactions on Knowledge and data Engineering 11, 1 (1999), 94--107.Google ScholarDigital Library
- Shipra Agrawal and Navin Goyal. 2013. Further optimal regret bounds for thompson sampling. In Artificial intelligence and statistics. 99--107.Google Scholar
- Susanne Albers. 1996. Competitive online algorithms. Citeseer.Google Scholar
- Noga Alon and Joel H Spencer. 2004. The probabilistic method. John Wiley & Sons.Google Scholar
- EC Amazon. 2015. Amazon web services. Available in: http://aws. amazon. com/es/ec2/(November 2012) (2015).Google Scholar
- Daniel Berend and Aryeh Kontorovich. 2013. A sharp estimate of the binomial mean absolute deviation with applications. Statistics & Probability Letters 83, 4 (2013), 1254--1259.Google ScholarCross Ref
- Daniel S Berger, Nathan Beckmann, and Mor Harchol-Balter. 2018. Practical bounds on optimal caching with variable object sizes. Proceedings of the ACM on Measurement and Analysis of Computing Systems 2, 2 (2018), 1--38. Proc. ACM Meas. Anal. Comput. Syst., Vol. 4, No. 2, Article 25. Publication date: June 2020. 25:28 Rajarshi Bhattacharjee, Subhankar Banerjee, and Abhishek SinhaGoogle ScholarDigital Library
- Dimitri P Bertsekas and Athena Scientific. 2015. Convex optimization algorithms. Athena Scientific Belmont.Google Scholar
- Timm Böttger, Felix Cuadrado, Gareth Tyson, Ignacio Castro, and Steve Uhlig. 2018. Open connect everywhere: A glimpse at the internet ecosystem through the lens of the netflix cdn. ACM SIGCOMM Computer Communication Review 48, 1 (2018), 28--34.Google ScholarDigital Library
- Lee Breslau, Pei Cao, Li Fan, Graham Phillips, Scott Shenker, et al. 1999. Web caching and Zipf-like distributions: Evidence and implications. In Ieee Infocom, Vol. 1. INSTITUTE OF ELECTRICAL ENGINEERS INC (IEEE), 126--134.Google ScholarCross Ref
- Nicolo Cesa-Bianchi and Gábor Lugosi. 2006. Prediction, learning, and games. Cambridge university press.Google Scholar
- Jacob Chakareski. 2017. VR/AR immersive communication: Caching, edge computing, and transmission trade-offs. In Proceedings of the Workshop on Virtual Reality and Augmented Reality Network. ACM, 36--41.Google ScholarDigital Library
- David Chappell et al. 2010. Introducing the windows azure platform. David Chappell & Associates White Paper (2010).Google Scholar
- Ludmila Cherkasova and Gianfranco Ciardo. 2001. Role of aging, frequency, and size in web cache replacement policies. In International Conference on High-Performance Computing and Networking. Springer, 114--123.Google ScholarCross Ref
- Alon Cohen and Tamir Hazan. 2015. Following the perturbed leader for online structured learning. In International Conference on Machine Learning. 1034--1042.Google Scholar
- Richard Combes, Stefan Magureanu, Alexandre Proutiere, and Cyrille Laroche. 2015. Learning to rank: Regret lower bounds and efficient algorithms. ACM SIGMETRICS Performance Evaluation Review 43, 1 (2015), 231--244.Google ScholarDigital Library
- Asit Dan and Don Towsley. 1990. An approximate analysis of the LRU and FIFO buffer replacement schemes. Vol. 18. ACM.Google Scholar
- Alexandros G Dimakis, P Brighten Godfrey, Yunnan Wu, Martin J Wainwright, and Kannan Ramchandran. 2010. Network coding for distributed storage systems. IEEE transactions on information theory 56, 9 (2010), 4539--4551.Google Scholar
- Gil Einziger, Roy Friedman, and Ben Manes. 2017. Tinylfu: A highly efficient cache admission policy. ACM Transactions on Storage (ToS) 13, 4 (2017), 1--31.Google ScholarDigital Library
- D. H. Finstad, H. K. Stensland, H. Espeland, and P. Halvorsen. 2011. Improved Multi-Rate Video Encoding. In 2011 IEEE International Symposium on Multimedia. 293--300. https://doi.org/10.1109/ISM.2011.53Google Scholar
- Philippe Flajolet, Daniele Gardy, and Loÿs Thimonier. 1992. Birthday paradox, coupon collectors, caching algorithms and self-organizing search. Discrete Applied Mathematics 39, 3 (1992), 207--229.Google ScholarDigital Library
- Gaston H Gonnet. 1981. Expected length of the longest probe sequence in hash code searching. Journal of the ACM (JACM) 28, 2 (1981), 289--304.Google ScholarDigital Library
- Steffen Grünewälder, Jean-Yves Audibert, Manfred Opper, and John Shawe-Taylor. 2010. Regret bounds for Gaussian process bandit problems. In Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics. 273--280.Google Scholar
- Chuang Gu, Chun-Wei Chan, William Chen, Stacey Spears, ThomasWHolcomb, Chih-Lung Lin, and Sanjeev Mehrotra. 2013. Multiple bit rate video encoding using variable bit rate and dynamic resolution for adaptive video streaming. US Patent 8,396,114.Google Scholar
- Ori Gurel-Gurevich. [n.d.]. Expectation of square root of binomial r.v. MathOverflow. arXiv:https://mathoverflow.net/q/121424 https://mathoverflow.net/q/121424 URL:https://mathoverflow.net/q/121424 (version: 2013-02--10).Google Scholar
- F Maxwell Harper and Joseph A Konstan. 2015. The movielens datasets: History and context. Acm transactions on interactive intelligent systems (tiis) 5, 4 (2015), 1--19.Google Scholar
- Elad Hazan, Amit Agarwal, and Satyen Kale. 2007. Logarithmic regret algorithms for online convex optimization. Machine Learning 69, 2--3 (2007), 169--192.Google ScholarDigital Library
- Elad Hazan and Sanjeev Arora. 2006. Efficient algorithms for online convex optimization and their applications. Princeton University Princeton.Google Scholar
- Elad Hazan and Satyen Kale. 2014. Beyond the regret minimization barrier: optimal algorithms for stochastic stronglyconvex optimization. The Journal of Machine Learning Research 15, 1 (2014), 2489--2512.Google ScholarDigital Library
- Thomas W Holcomb, Shankar Regunathan, Chih-lung Bruce Lin, and Sridhar Srinivasan. 2008. Multi-resolution video coding and decoding. US Patent 7,379,496.Google Scholar
- Predrag R Jelenkovi? and Xiaozhu Kang. 2008. Characterizing the miss sequence of the LRU cache. ACM SIGMETRICS Performance Evaluation Review 36, 2 (2008), 119--121.Google ScholarDigital Library
- M. Ji, G. Caire, and A. F. Molisch. 2016. Fundamental Limits of Caching in Wireless D2D Networks. IEEE Transactions on Information Theory 62, 2 (Feb 2016), 849--869. https://doi.org/10.1109/TIT.2015.2504556Google ScholarCross Ref
- M. Ji, M. F. Wong, A. M. Tulino, J. Llorca, G. Caire, M. Effros, and M. Langberg. 2015. On the fundamental limits of caching in combination networks. In 2015 IEEE 16th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC). 695--699. https://doi.org/10.1109/SPAWC.2015.7227127Google ScholarCross Ref
- Valentin Fedorovich Kolchin, Boris Aleksandrovich Sevastyanov, and Vladimir Pavlovich Chistyakov. 1978. Random allocations. (1978). Proc. ACM Meas. Anal. Comput. Syst., Vol. 4, No. 2, Article 25. Publication date: June 2020. Fundamental Limits on the Regret of Online Network-Caching 25:29Google Scholar
- Branislav Kveton, Zheng Wen, Azin Ashkan, and Csaba Szepesvari. 2015. Tight regret bounds for stochastic combinatorial semi-bandits. In Artificial Intelligence and Statistics. 535--543.Google Scholar
- Donghee Lee, Jongmoo Choi, Jong-Hun Kim, Sam H Noh, Sang Lyul Min, Yookun Cho, and Chong-Sang Kim. 1999. On the existence of a spectrum of policies that subsumes the least recently used (LRU) and least frequently used (LFU) policies.. In SIGMETRICS, Vol. 99. Citeseer, 1--4.Google ScholarDigital Library
- Boxi Liu, Konstantinos Poularakis, Leandros Tassiulas, and Tao Jiang. 2019. Joint Caching and Routing in Congestible Networks of Arbitrary Topology. IEEE Internet of Things Journal (2019).Google Scholar
- M Luby, A Shokrollahi, M Watson, T Stockhammer, and L Minder. 2011. RaptorQ forward error correction scheme for object delivery?rfc 6330. IETF Request For Comments (2011).Google Scholar
- Thodoris Lykouris and Sergei Vassilvitskii. 2018. Competitive caching with machine learned advice. arXiv preprint arXiv:1802.05399 (2018).Google Scholar
- Mohammad Ali Maddah-Ali and Urs Niesen. 2014. Fundamental limits of caching. IEEE Transactions on Information Theory 60, 5 (2014), 2856--2867.Google ScholarCross Ref
- M. A. Maddah-Ali and U. Niesen. 2016. Coding for caching: fundamental limits and practical challenges. IEEE Communications Magazine 54, 8 (August 2016), 23--29. https://doi.org/10.1109/MCOM.2016.7537173Google ScholarDigital Library
- Bruce M Maggs and Ramesh K Sitaraman. 2015. Algorithmic nuggets in content delivery. ACM SIGCOMM Computer Communication Review 45, 3 (2015), 52-66.Google ScholarDigital Library
- Stefan Magureanu, Richard Combes, and Alexandre Proutière. 2014. Lipschitz Bandits: Regret Lower Bound and Optimal Algorithms. In Proceedings of The 27th Conference on Learning Theory, COLT 2014, Barcelona, Spain, June 13--15, 2014. 975--999. http://proceedings.mlr.press/v35/magureanu14.htmlGoogle Scholar
- Michael Mitzenmacher and Eli Upfal. 2017. Probability and computing: Randomization and probabilistic techniques in algorithms and data analysis. Cambridge university press.Google Scholar
- Michael David Mitzenmacher. 1996. The Power of Two Choices in Randomized Load Balancing. Ph.D. Dissertation. UNIVERSITY of CALIFORNIA at BERKELEY.Google Scholar
- Erik Nygren, Ramesh K Sitaraman, and Jennifer Sun. 2010. The akamai network: a platform for high-performance internet applications. ACM SIGOPS Operating Systems Review 44, 3 (2010), 2-19.Google ScholarDigital Library
- Georgios Paschos, George Iosifidis, and Giuseppe Caire. 2019. Cache Optimization Models and Algorithms. arXiv preprint arXiv:1912.12339 (2019).Google Scholar
- Georgios S Paschos, Apostolos Destounis, Luigi Vigneri, and George Iosifidis. 2019. Learning to Cache With No Regrets. In IEEE INFOCOM 2019-IEEE Conference on Computer Communications. IEEE, 235--243.Google ScholarDigital Library
- Ramtin Pedarsani, Mohammad Ali Maddah-Ali, and Urs Niesen. 2016. Online coded caching. IEEE/ACM Transactions on Networking (TON) 24, 2 (2016), 836--845.Google ScholarDigital Library
- Martin Raab and Angelika Steger. 1998. "Balls into bins"Ð A simple and tight analysis. In International Workshop on Randomization and Approximation Techniques in Computer Science. Springer, 159--170.Google ScholarCross Ref
- Amr Rizk, Michael Zink, and Ramesh Sitaraman. 2017. Model-based design and analysis of cache hierarchies. In 2017 IFIP Networking Conference (IFIP Networking) and Workshops. IEEE, 1--9.Google ScholarCross Ref
- Herbert Robbins. 1955. A remark on Stirling's formula. The American mathematical monthly 62, 1 (1955), 26--29.Google Scholar
- Elisha J Rosensweig, Daniel S Menasche, and Jim Kurose. 2013. On the steady-state of cache networks. In 2013 Proceedings IEEE INFOCOM. IEEE, 863--871.Google ScholarCross Ref
- Walter Rudin et al. 1964. Principles of mathematical analysis. Vol. 3. McGraw-hill New York.Google Scholar
- D. N. Serpanos, G. Karakostas, and W. H. Wolf. 2000. Effective caching of Web objects using Zipf's law. In 2000 IEEE International Conference on Multimedia and Expo. ICME2000. Proceedings. Latest Advances in the Fast Changing World of Multimedia (Cat. No.00TH8532), Vol. 2. 727--730 vol.2. https://doi.org/10.1109/ICME.2000.871464Google Scholar
- Shai Shalev-Shwartz et al. 2012. Online learning and online convex optimization. Foundations and Trends® in Machine Learning 4, 2 (2012), 107--194.Google Scholar
- Karthikeyan Shanmugam, Negin Golrezaei, Alexandros G Dimakis, Andreas F Molisch, and Giuseppe Caire. 2013. Femtocaching: Wireless content delivery through distributed caching helpers. IEEE Transactions on Information Theory 59, 12 (2013), 8402-8413.Google ScholarDigital Library
- Amin Shokrollahi. 2006. Raptor codes. IEEE/ACM Transactions on Networking (TON) 14, SI (2006), 2551--2567.Google Scholar
- Konstantin Shvachko, Hairong Kuang, Sanjay Radia, Robert Chansler, et al. 2010. The hadoop distributed file system.. In MSST, Vol. 10. 1--10.Google Scholar
- Abraham Silberschatz, Peter Baer Galvin, and Greg Gagne. 2006. Operating system principles. John Wiley & Sons.Google Scholar
- Stefano Traverso, Mohamed Ahmed, Michele Garetto, Paolo Giaccone, Emilio Leonardi, and Saverio Niccolini. 2015. Unravelling the impact of temporal and geographical locality in content caching systems. IEEE Transactions on Multimedia 17, 10 (2015), 1839--1854.Google ScholarDigital Library
- Benjamin Van Roy. 2007. A short proof of optimality for the MIN cache replacement algorithm. Information processing letters 102, 2--3 (2007), 72-73. Proc. ACM Meas. Anal. Comput. Syst., Vol. 4, No. 2, Article 25. Publication date: June 2020. 25:30 Rajarshi Bhattacharjee, Subhankar Banerjee, and Abhishek SinhaGoogle Scholar
- Sarut Vanichpun and Armand M Makowski. 2004. The output of a cache under the independent reference model: where did the locality of reference go?. In Proceedings of the joint international conference on Measurement and modeling of computer systems. 295--306.Google ScholarDigital Library
- Jinesh Varia, Sajee Mathew, et al. 2014. Overview of amazon web services. Amazon Web Services (2014), 1--22.Google Scholar
- Chuan Wu and Baochun Li. 2007. rStream: resilient and optimal peer-to-peer streaming with rateless codes. IEEE Transactions on Parallel and Distributed Systems 19, 1 (2007), 77--92.Google Scholar
Index Terms
- Fundamental Limits on the Regret of Online Network-Caching
Recommendations
Optimistic No-regret Algorithms for Discrete Caching
POMACSWe take a systematic look at the problem of storing whole files in a cache with limited capacity in the context of optimistic learning, where the caching policy has access to a prediction oracle (provided by, e.g., a Neural Network). The successive file ...
Fundamental Limits on the Regret of Online Network-Caching
SIGMETRICS '20: Abstracts of the 2020 SIGMETRICS/Performance Joint International Conference on Measurement and Modeling of Computer SystemsOptimal caching of files in a content distribution network (CDN) is a problem of fundamental and growing commercial interest. Although many different caching algorithms are in use today, the fundamental performance limits of the network caching ...
Fundamental Limits on the Regret of Online Network-Caching
Optimal caching of files in a content distribution network (CDN) is a problem of fundamental and growing commercial interest. Although many different caching algorithms are in use today, the fundamental performance limits of the network caching ...
Comments