Caching In Content Delivery Networks Research Articles

SLAP: Segmented Reuse-Time-Label Based Admission Policy for Content Delivery Network Caching

‘‘Learned” admission policies have shown promise in improving Content Delivery Network (CDN) cache performance and lowering operational costs. Unfortunately, existing learned policies are optimized with a few fixed cache sizes while in reality, cache sizes often vary over time in an unpredictable manner. As a result, existing solutions cannot provide consistent benefits in production settings. We present SLAP , a learned CDN cache admission approach based on segmented object reuse time prediction. SLAP predicts an object’s reuse time range using the Long-Short-Term-Memory model and admits objects that will be reused (before eviction) given the current cache size. SLAP decouples model training from cache size, allowing it to adapt to arbitrary sizes. The key to our solution is a novel segmented labeling scheme that makes SLAP without requiring precise prediction on object reuse time. To further make SLAP a practical and efficient solution, we propose aggressive reusing of computation and training on sampled traces to optimize model training, and a specialized predictor architecture that overlaps prediction computation with miss object fetching to optimize model inference. Our experiments using production CDN traces show that SLAP achieves significantly lower write traffic (38%-59%), longer SSDs lifetime (104%-178%), a consistently higher hit rate (3.2%-11.7%), and requires no effort to adapt to changing cache sizes, outperforming existing policies.

Video Quality and Popularity-Aware Video Caching in Content Delivery Networks

Video Quality and Popularity-Aware Video Caching in Content Delivery Networks

Novel Architecture of Content Delivery Network Based on Big Data for Power Saving

This paper proposes an architecture of content delivery network (CDN) based on big data for power saving. There are two types of video content: hot content and cold content. When video content is accessed frequently, it is called hot content. Conversely, when video content is accessed infrequently, it is called cold content. In CDN, there is an origin server and a CDN cache server. A CDN cache server has a replicated content and provides its content to the end users nearby. Therefore, the user can receive the requested content from the closest proximity for fast content. The proposed architecture in this paper powers off the cold content server in CDN cache server when the number of cold content requests decreases. Hence, the proposed architecture for content delivery services based on power saving is expected to be useful for providing multimedia streaming services with low power consumption for content providers.

RL-Cache: Learning-Based Cache Admission for Content Delivery

Content delivery networks (CDNs) distribute much of the Internet content by caching and serving the objects requested by users. A major goal of a CDN is to maximize the hit rates of its caches, thereby enabling faster content downloads to the users. Content caching involves two components: an admission algorithm to decide whether to cache an object and an eviction algorithm to determine which object to evict from the cache when it is full. In this paper, we focus on cache admission and propose a novel algorithm called RL-Cache that uses model-free reinforcement learning (RL) to decide whether or not to admit a requested object into the CDN's cache. Unlike prior approaches that use a small set of criteria for decision making, RL-Cache weights a large set of features that include the object size, recency, and frequency of access. We develop a publicly available implementation of RL-Cache and perform an evaluation using production traces for the image, video, and web traffic classes from Akamai's CDN. The evaluation shows that RL-Cache improves the hit rate in comparison with the state of the art and imposes only a modest resource overhead on the CDN servers. Further, RL-Cache is robust enough that it can be trained in one location and executed on request traces of the same or different traffic classes in other locations of the same geographic region. The paper also reports extensive analyses of the RL-Cache sensitivity to its features and hyperparameter values. The analyses validate the made design choices and reveal interesting insights into the RL-Cache behavior.

Reproducing Popularity Distribution of YouTube Videos

To provide video streaming of user-generated contents (UGCs) with high quality and at low cost by maximizing the effect of content delivery network (CDN), CDN providers are required to adequately design CDN cache servers by accurately estimating the UGC view-count distribution. To achieve this goal in a practical time frame, we need to construct a simple time-series model that captures the transition of UGC popularity. Therefore, in this paper, we first analyze the daily view count (DVC) of YouTube videos over nine months and find that the DVC of YouTube videos obeys a lognormal distribution. As a simple time-series model of the DVC of each YouTube video, we propose the grouped MPP (gMPP), extending the multiplicative process (MPP) which is widely known as a simple time-series model generating a lognormal distribution. We also propose reproducing the DVC distribution of YouTube videos by using a superposed gMPP (SgMPP) that aggregates multiple gMPPs. The SgMPP can accurately reproduce the DVC distribution of YouTube videos with a low computational overhead, so we can expect to use the SgMPP as the input for computer simulations for designing various network components that require the popularity distribution of UGC, e.g., cache capacities. Through numerical evaluation, we confirm that we can adequately design the storage capacity of a cache server with the average error rate of several percent against the target cache hit ratio.

Popularity prediction–based caching in content delivery networks

In content delivery networks (CDNs), caches are resources that must be allocated. For that purpose, videos’ popularity knowledge helps to make efficient decisions about which videos should be cached. Thus, we must be able to anticipate future needs in terms of requested videos. To do this, we rely on the requests history. This paper focuses on predicting the videos’ popularity: the daily number of requests. For that purpose, we propose a two-level prediction approach. At the first level, the experts compute the videos’ popularity, each expert using its own prediction method with its own parameters. At the second level, the forecasters select the best experts and build a prediction based on the predictions provided by these experts. The prediction accuracy is evaluated by a loss function as the discrepancy between the prediction value and the real number of requests. We use real traces extracted from YouTube to compare different prediction methods and determine the best parameter tuning for experts and forecasters. The goal is to find the best trade-off between complexity and accuracy of the prediction methods used. Finally, we apply these prediction methods to caching. Prediction methods are compared in terms of cache hit ratio and update ratio. The gain brought by this two-level prediction approach is compared with that obtained by a single prediction level. The results show that the choice of a two-level prediction approach is justified.

A user‐centric cooperative edge caching scheme for minimizing delay in 5G content delivery networks

AbstractThis paper investigates the problem of user‐centric cooperative edge caching in content delivery networks to leverage service provisioning at the network edge and to improve the quality of experience by minimizing the end‐to‐end delay. By taking advantage of the major characteristics in fifth‐generation networks, users can access contents not only from the nearest small base station (SBS) but also from other SBSs in the vicinity that have the requested precached contents. In the proposed scheme, a user‐centric delivery approach is considered in such a way that the base station can respond to the user request as long as it has enough resources. To this end, a caching algorithm is introduced whereby a group of SBSs cooperatively share storage and decide on the caching policy together aiming to cache as much contents as possible under the capacity constraint. Moreover, a modified matching theory is used for content delivery taking the bandwidth requirements into account. Simulation results show that the proposed scheme can achieve 99.5% local serve ratio when the SBS has a caching capacity of 40% the overall file size, and 20 available communication channels for content distribution.

A data-driven approach of performance evaluation for cache server groups in content delivery network

In industry, Content Delivery Network (CDN) service providers are increasingly using data-driven mechanisms to build the performance models of the service-providing systems. Building a model to accurately describe the performance of the existing infrastructure is very crucial to make resource management decisions. Conventional approaches that use hand-tuned parameters or linear models have their drawbacks. Recently, data-driven paradigm has been shown to greatly outperform traditional methods in modeling complex systems. We design a data-driven approach to building a reasonable and feasible performance model for CDN cache server groups. We use deep LSTM auto-encoder to capture the temporal structures from the high-dimensional monitoring data, and use a deep neural network to predict the reach rate which is a client QoS measurement from the CDN service providers’ perspective. The experimental results have shown that our model is able to outperform state-of-the-art models.

A Strategy of CDN Traffic Optimization Based on the Technology of SDN

Modern Content Delivery Network(CDN)[1,2] by caching and scheduling strategy to reduce user-perceived latency, because CDN can schedule user to the CDN cache node which closest to. At the same time, this strategy can reduce the data flow which has to be sent back to the source station, and decrease the calculation pressure of source station. For some content which is dynamic and cannot be cached, the current acceleration approach is to create a GRE tunnel between the CDN nodes, using the static route scheduling dynamic data flow, improve the speed of data flow which back to the source station. It has been a recent trend of CDN to separate the control flow and data flow through technical, in order to change traditional dynamic/static content acceleration model, to optimize the CDN traffic. The core idea of Software Defined Network(SDN)[3,4]is through the network control plane and data plane separation equipment, so as to realize the flexible control of network traffic. SDN technology is used to transform the CDN, Make the CDN have an intelligent traffic scheduling environment which can focus on monitoring network and scheduling flow path real-time According to the different business. Ultimate aim is to avoid local congestion, optimize the CDN traffic, and improve the quality of the CDN service effect.

Reducing Latency Through Page-aware Management of Web Objects by Content Delivery Networks

As popular web sites turn to content delivery networks (CDNs) for full-site delivery, there is an opportunity to improve the end-user experience by optimizing the delivery of entire web pages, rather than just individual objects. In particular, this paper explores page-structure-aware strategies for placing objects in CDN cache hierarchies. The key idea is that the objects in a web page that have the largest impact on page latency should be served out of the closest or fastest caches in the hierarchy. We present schemes for identifying these objects and develop mechanisms to ensure that they are served with higher priority by the CDN, while balancing traditional CDN concerns such as optimizing the delivery of popular objects and minimizing bandwidth costs. To establish a baseline for evaluating improvements in page latencies, we collect and analyze publicly visible HTTP headers that reveal the distribution of objects among the various levels of a major CDN's cache hierarchy. Through extensive experiments on 83 real-world web pages, we show that latency reductions of over 100 ms can be obtained for 30% of the popular pages, with even larger reductions for the less popular pages. Using anonymized server logs provided by the CDN, we show the feasibility of reducing capacity and staleness misses of critical objects by 60% with minimal increase in overall miss rates, and bandwidth overheads of under 0.02%.

Revisiting caching in content delivery networks

Content Delivery Networks (CDNs) differ from other caching systems in terms of both workload characteristics and performance metrics. However, there has been little prior work on large-scale measurement and characterization of content requests and caching performance in CDNs. For workload characteristics, CDNs deal with extremely large content volume, high content diversity, and strong temporal dynamics. For performance metrics, other than hit ratio, CDNs also need to minimize the disk operations and the volume of traffic from origin servers. In this paper, we conduct a large-scale measurement study to characterize the content request patterns using real-world data from a commercial CDN provider.