Analysis of tandem PIT and CS with non-zero download delay

Dai, Honghua; Liu, Bin; Yuan, Haowei; Crowley, Patrick; Lü, Jian

doi:10.1109/infocom.2017.8057191

Cited by 9 publications

(9 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recall that this assumption has been shown to produce significant errors (even in the zero object fetch delay regime) [7]. Further work that considers the download delay time for objects in single LRU caches includes [26,27], where the authors propose variants of the LRU algorithm that work better under non-zero delay by jointly analyzing single LRU caching and pending interest table (PIT) that contain object requests that are not yet fetched. The authors of [28] consider the object download delay and propose an algorithm based on q-LRU to optimize the average latency based on the popularity of the object and its size.…”

Section: Analysis Of Ttl Cache Hierarchiesmentioning

confidence: 99%

“…How our contributions differ from [7,10,[26][27][28]: While [10] analyzes cache hierarchies under non-zero object fetch delays using approximations, our work differs significantly as we exactly analyze the caching hierarchies under non-zero delay. In Sect.5 and in the evaluations in Sect.…”

Section: Analysis Of Ttl Cache Hierarchiesmentioning

confidence: 99%

“…Note that the work in [10] is exact in case of a single cache. Further, our work is different from [26][27][28] due to the same technical reasons, as well as, the fact that we consider a fundamentally different caching algorithm (TTL instead of LRU). Our contributions also differ from the work in [7] which assumes zero object fetch delay and focuses on the derivation of cache performance metrics.…”

Section: Analysis Of Ttl Cache Hierarchiesmentioning

confidence: 99%

See 2 more Smart Citations

On the Impact of Network Delays on Time-to-Live Caching

El‐Sayed¹,

Rizk²

2022

Preprint

View full text Add to dashboard Cite

We consider Time-to-Live (TTL) caches that tag every object in cache with a specific (and possibly renewable) expiration time. State-of-the-art models for TTL caches assume zero object fetch delay, i.e., the time required to fetch a requested object that is not in cache from a different cache or the origin server. Particularly, in cache hierarchies this delay has a significant impact on performance metrics such as the object hit probability. Recent work suggests that the impact of the object fetch delay on the cache performance will continue to increase due to the scaling mismatch between shrinking inter-request times (due to higher data center link rates) in contrast to processing and memory access times.In this paper, we analyze tree-based cache hierarchies with random object fetch delays and provide an exact analysis of the corresponding object hit probability. Our analysis allows understanding the impact of random delays and TTLs on cache metrics for a wide class of request stream models characterized through Markov arrival processes. This is expressed through a metric that we denote delay impairment of the hit probability. In addition, we analyze and extend state-of-the-art approximations of the hit probability to take the delay into account. We provide numerical and trace-based simulation-based evaluation results showing that larger TTLs do not efficiently compensate the detrimental effect of object fetch delays. Our evaluations also show that unlike our exact model the state-of-the-art approximations do not capture the impact of the object fetch delay well especially for cache hierarchies. Surprisingly, we show that the impact of this delay on the hit probability is not monotonic but depends on the request stream properties, as well as, the TTL.

show abstract

Section: Analysis Of Ttl Cache Hierarchiesmentioning

confidence: 99%

Section: Analysis Of Ttl Cache Hierarchiesmentioning

confidence: 99%

Section: Analysis Of Ttl Cache Hierarchiesmentioning

confidence: 99%

See 1 more Smart Citation

On the Impact of Network Delays on Time-to-Live Caching

El‐Sayed¹,

Rizk²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In recent years, the realistic scenarios with request aggregation have been considered. It has been theoretically proved that the random process of the request arrival distribution and the content arrival distribution became different due to the request aggregation mechanism [20], [21]. The cache hit rate, average response time of requests and the cache space size distribution occupied by the pending interest table (PIT) were analyzed by using the cache strategy with fixed content sojourn time [30].…”

Section: Related Workmentioning

confidence: 99%

“…Their formulated optimization problems are able to approximate to the actual system performances with low content request intensity. However, in the scenarios of massive content requests, the performances become distinctive due to the inherent request aggregation mechanism in CCN [20], [21].…”

Section: Introductionmentioning

confidence: 99%

Proactive Edge Caching in Content-Centric Networks With Massive Dynamic Content Requests

Feng

Shan

et al. 2020

IEEE Access

View full text Add to dashboard Cite

Edge computing is a promising infrastructure evolution to reduce traffic loads and support low-latency communications. Furthermore, content-centric networks provide a natural solution to cache contents at edge nodes. However, it is a challenge for edge nodes to handle massive and highly dynamic content requests by users, and if without an efficient content caching strategy, the edge nodes will encounter high traffic load and latency due to increasing retrieval from content providers. This paper formulates a proactive edge caching problem to minimize the content retrieval cost at edge nodes. We exploit the inherent content caching and request aggregation mechanism in the content-centric networks to jointly minimize traffic load and content retrieval delay cost generated by the massive and dynamic content requests. We develop a Q-learning algorithm, which is an online optimal caching strategy, as it is adaptable to dynamic content popularity and content request intensity, and derive the long-term minimization of the content retrieval cost. Simulation results illustrate that the proposed algorithm can achieve a lower content retrieval cost compared with several baseline caching schemes. INDEX TERMS Content-centric networks, dynamic content requests, edge computing, proactive caching, Q-learning. XIAOGENG XU (Student Member, IEEE) received the B.S. degree in communications engineering from the Beijing University of Technology, China, in 2005, the M.S. degree in digital communication systems and technology from the Chalmers University of Technology, Sweden, in 2007. She is currently pursuing the Ph.D. degree in communication and information systems from the Beijing University of Posts and Telecommunications. Her current researches focus on caching performance analysis and optimization in content-centric networks. CHUNYAN FENG (Senior Member, IEEE) received the B.S. degree in communications engineering and the M.S. and Ph.D. degrees in communication and information systems from the Beijing University of Posts and Telecommunications (BUPT), Beijing, China. She is currently a Professor with the School of Information and Communication Engineering, BUPT. Her research interests are in the areas of broadband networks and wireless communication systems. Her current research focuses on cognitive radio and green wireless communications.

show abstract