Information-theoretic caching

Wang, Chien-Yi; Lim, Sung Hoon; Gastpar, Michael

doi:10.1109/isit.2015.7282761

Cited by 11 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several coded caching schemes have been proposed since then [11]- [16]. The caching problem has also been extended in various directions, including decentralized caching [17], online caching [18], caching with nonuniform demands [19]- [22], hierarchical caching [23]- [25], device-to-device caching [26], cache-aided interference channels [27]- [30], caching on file selection networks [31]- [33], caching on broadcast channels [34]- [37], and caching for channels with delayed feedback with channel state information [38], [39]. The same idea is also useful in the context of distributed computing, in order to take advantage of extra computation to reduce the communication load [40]- [44].…”

Section: Introductionmentioning

confidence: 99%

The exact rate-memory tradeoff for caching with uncoded prefetching

Maddah-Ali

Avestimehr

2017

2017 IEEE International Symposium on Information Theory (ISIT)

156

441

View full text Add to dashboard Cite

Abstract-We consider a basic cache network, in which a single server is connected to multiple users via a shared bottleneck link. The server has a database of files (content). Each user has an isolated memory that can be used to cache content in a prefetching phase. In a following delivery phase, each user requests a file from the database, and the server needs to deliver users' demands as efficiently as possible by taking into account their cache contents. We focus on an important and commonly used class of prefetching schemes, where the caches are filled with uncoded data. We provide the exact characterization of the rate-memory tradeoff for this problem, by deriving both the minimum average rate (for a uniform file popularity) and the minimum peak rate required on the bottleneck link for a given cache size available at each user. In particular, we propose a novel caching scheme, which strictly improves the state of the art by exploiting commonality among user demands. We then demonstrate the exact optimality of our proposed scheme through a matching converse, by dividing the set of all demands into types, and showing that the placement phase in the proposed caching scheme is universally optimal for all types. Using these techniques, we also fully characterize the rate-memory tradeoff for a decentralized setting, in which users fill out their cache content without any coordination.

show abstract

Section: Introductionmentioning

confidence: 99%

The exact rate-memory tradeoff for caching with uncoded prefetching

Maddah-Ali

Avestimehr

2017

2017 IEEE International Symposium on Information Theory (ISIT)

156

441

View full text Add to dashboard Cite

show abstract

“…The standard Gray-Wyner network [11], which consists of encoder 1 and decoders 1, 2, is closely related to the standard (static) caching problem [4] in the sense that the common link (M 0 ) can be seen as the cache and private links can be seen as the updates for the corresponding requests. Similarly, the successive Gray-Wyner network captures the dynamic nature of the content.…”

Section: Successive Gray-wyner Networkmentioning

confidence: 99%

“…Finally, in the delivery phase, the remaining part of the requested files is delivered. Taking an information theoretical approach, [4], [10] formulated the single user caching problem through its similarity to Gray-Wyner network [11] and discussed the optimal caching strategy. Extension of the theoretical analysis to multi-user caching problems can be found in [2], [5], [10], [12].…”

Section: Introductionmentioning

confidence: 99%

“…What would be the optimal caching strategy in such a setup to minimize the expected delivery rate? Following similar steps to [4], we approach this question from an information theoretical perspective and construct a successive Gray-Wyner network to exploit the dynamic nature of the content files. We first characterize the optimal rate region for the successive Gray-Wyner network (cf.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Successive Refinement to Caching for Dynamic Content

Sen

Gastpar

2019

2019 IEEE International Symposium on Information Theory (ISIT)

Self Cite

View full text Add to dashboard Cite

To reduce the network load during peak hours, servers deliver partial data to users during the off-peak time of the network before the actual requests are known, which is known as caching. This paper studies a single user caching problem in which the file contents are subject to dynamic modifications with respect to a certain probability distribution. To cope with the dynamical nature of the file contents, a successive refinement approach to caching is presented: partial information of the original data is cached first and then if there is a modification, a refinement to the previously cached data is delivered to the user. Given a fixed cache memory, there is a tension between the rates of two cache descriptions. The problem of optimal caching strategies is formulated through a successive Gray-Wyner network, the optimal rate region of which is characterized. Some lower and upper bounds on the performance of optimal caching strategies are developed and shown to actually yield closed form solutions for certain classes of file contents.

show abstract

“…In this paper, we study this interpretation of caching applied to Gaussians, similar to the approach of [3], [4] for discrete sources. As a start, we assume the requests are uniformly distributed; the focus of our study is to achieve efficient caching through exploiting correlation in the data streams.…”

Section: Introductionmentioning

confidence: 99%

Caching Gaussians: Minimizing total correlation on the Gray-Wyner network

Veld

Gastpar

2016

2016 Annual Conference on Information Science and Systems (CISS)

View full text Add to dashboard Cite

We study a caching problem that resembles a lossy Gray-Wyner network: A source produces vector samples from a Gaussian distribution, but the user is interested in the samples of only one component. The encoder first sends a cache message without any knowledge of the user's preference. Upon learning her request, a second message is provided in the update phase so as to attain the desired fidelity on that component.The cache is efficient if it exploits as much of the correlation in the source as possible, which connects to the notions of Wyner's common information (for high cache rates) and Watanabe's total correlation (for low cache rates). For the former, we extend known results for 2 Gaussians to multivariates by showing that common information is a simple linear program, which can be solved analytically for circulant correlation matrices.Total correlation in a Gaussian setting is less well-studied. We show that for bivariates and using Gaussian auxiliaries it is captured in the dominant eigenvalue of the correlation matrix. For multivariates the problem is a more difficult optimization over a non-convex domain, but we conjecture that circulant matrices may again be analytically solvable.

show abstract

Information-theoretic caching

Cited by 11 publications

References 14 publications

The exact rate-memory tradeoff for caching with uncoded prefetching

The exact rate-memory tradeoff for caching with uncoded prefetching

Successive Refinement to Caching for Dynamic Content

Caching Gaussians: Minimizing total correlation on the Gray-Wyner network

Contact Info

Product

Resources

About