The Experts below are selected from a list of 13092 Experts worldwide ranked by ideXlab platform
Anirban Mahanti - One of the best experts on this subject based on the ideXlab platform.
-
youtube Traffic Characterization a view from the edge
Internet Measurement Conference, 2007Co-Authors: Phillipa Gill, Martin Arlitt, Anirban MahantiAbstract:This paper presents a Traffic Characterization study of the popular video sharing service, YouTube. Over a three month period we observed almost 25 million transactions between users on an edge network and YouTube, including more than 600,000 video downloads. We also monitored the globally popular videos over this period of time. In the paper we examine usage patterns, file properties, popularity and referencing characteristics, and transfer behaviors of YouTube, and compare them to traditional Web and media streaming workload characteristics. We conclude the paper with a discussion of the implications of the observed characteristics. For example, we find that as with the traditional Web, caching could improve the end user experience, reduce network bandwidth consumption, and reduce the load on YouTube's core server infrastructure. Unlike traditional Web caching, Web 2.0 provides additional meta-data that should be exploited to improve the effectiveness of strategies like caching.
-
Internet Measurement Comference - Youtube Traffic Characterization: a view from the edge
Proceedings of the 7th ACM SIGCOMM conference on Internet measurement - IMC '07, 2007Co-Authors: Phillipa Gill, Martin Arlitt, Anirban MahantiAbstract:This paper presents a Traffic Characterization study of the popular video sharing service, YouTube. Over a three month period we observed almost 25 million transactions between users on an edge network and YouTube, including more than 600,000 video downloads. We also monitored the globally popular videos over this period of time. In the paper we examine usage patterns, file properties, popularity and referencing characteristics, and transfer behaviors of YouTube, and compare them to traditional Web and media streaming workload characteristics. We conclude the paper with a discussion of the implications of the observed characteristics. For example, we find that as with the traditional Web, caching could improve the end user experience, reduce network bandwidth consumption, and reduce the load on YouTube's core server infrastructure. Unlike traditional Web caching, Web 2.0 provides additional meta-data that should be exploited to improve the effectiveness of strategies like caching.
Tsung Yuan Charles Tai - One of the best experts on this subject based on the ideXlab platform.
-
Network Traffic Characterization using token bucket model
Proceedings - IEEE INFOCOM, 1999Co-Authors: Puqi Perry Tang, Tsung Yuan Charles TaiAbstract:This paper investigated the problem of deriving token bucket parameters from the observed Traffic patterns of computer network flows in two cases. In the basic case, we identified the set of token bucket parameters for any given flow so that all the packets in the flow will be delivered immediately without incurring any delay or loss. Then, we extended the basic case by adding a queue to the token bucket model to temporarily store packets that cannot be delivered right away until enough tokens get accumulated in the bucket. The queue has the effect of smoothing the Traffic so that the resulting token bucket parameters are less demanding. It also makes the derived token bucket parameters less fluctuated over time. Relationship among the queue size and token bucket parameters for a given flow are analyzed rigorously. Queuing delay for each packet in the flow is calculated and used to describe the adjusted post-queuing Traffic pattern. Simple and efficient algorithms for the derivation of measurement-based Traffic specification (MBTS) are presented, along with empirical results. This MBTS technology alleviates the need for users to explicitly characterize the Traffic a priori in order to reserve network resource in an integrated services network
Nicolas Larrieu - One of the best experts on this subject based on the ideXlab platform.
-
HSNMC - Internet Traffic Characterization – An Analysis of Traffic Oscillations
Lecture Notes in Computer Science, 2004Co-Authors: Philippe Owezarski, Nicolas LarrieuAbstract:Internet Traffic has been changing a lot since few years in particular with the arrival of new P2P applications for exchanging audio files or movies and nowadays the knowledge we have on it is quite limited. Especially, new applications and new Traffic are creating a lot of troubles and performance issues. Based on some Traffic traces captured in the framework of the METROPOLIS network monitoring project, this paper exhibits the highly oscillating nature of Internet Traffic, thus explaining why it is almost impossible nowadays to guarantee a stable QoS in the Internet, and also that such oscillations provoke a huge decrease of the global network QoS and performance. This paper then demonstrates that Traffic oscillations can be characterized by the Hurst (LRD) parameter. In particular, this demonstration relies on a comparative study of Internet Traffic depending on the transport protocol used to generate it. It is then shown that using TFRC – a congestion control mechanism whose purpose deals with providing smooth sending rates for stream oriented applications – instead of TCP, makes Traffic oscillations and LRD almost disappear. This result, i.e. limiting as much as possible the oscillations of Traffic sources in the Internet, then gives research directions for future Internet protocols and architectures.
-
internet Traffic Characterization an analysis of Traffic oscillations
International Conference on Communications, 2004Co-Authors: Philippe Owezarski, Nicolas LarrieuAbstract:Internet Traffic has been changing a lot since few years in particular with the arrival of new P2P applications for exchanging audio files or movies and nowadays the knowledge we have on it is quite limited. Especially, new applications and new Traffic are creating a lot of troubles and performance issues. Based on some Traffic traces captured in the framework of the METROPOLIS network monitoring project, this paper exhibits the highly oscillating nature of Internet Traffic, thus explaining why it is almost impossible nowadays to guarantee a stable QoS in the Internet, and also that such oscillations provoke a huge decrease of the global network QoS and performance. This paper then demonstrates that Traffic oscillations can be characterized by the Hurst (LRD) parameter. In particular, this demonstration relies on a comparative study of Internet Traffic depending on the transport protocol used to generate it. It is then shown that using TFRC – a congestion control mechanism whose purpose deals with providing smooth sending rates for stream oriented applications – instead of TCP, makes Traffic oscillations and LRD almost disappear. This result, i.e. limiting as much as possible the oscillations of Traffic sources in the Internet, then gives research directions for future Internet protocols and architectures.
Phillipa Gill - One of the best experts on this subject based on the ideXlab platform.
-
youtube Traffic Characterization a view from the edge
Internet Measurement Conference, 2007Co-Authors: Phillipa Gill, Martin Arlitt, Anirban MahantiAbstract:This paper presents a Traffic Characterization study of the popular video sharing service, YouTube. Over a three month period we observed almost 25 million transactions between users on an edge network and YouTube, including more than 600,000 video downloads. We also monitored the globally popular videos over this period of time. In the paper we examine usage patterns, file properties, popularity and referencing characteristics, and transfer behaviors of YouTube, and compare them to traditional Web and media streaming workload characteristics. We conclude the paper with a discussion of the implications of the observed characteristics. For example, we find that as with the traditional Web, caching could improve the end user experience, reduce network bandwidth consumption, and reduce the load on YouTube's core server infrastructure. Unlike traditional Web caching, Web 2.0 provides additional meta-data that should be exploited to improve the effectiveness of strategies like caching.
-
Internet Measurement Comference - Youtube Traffic Characterization: a view from the edge
Proceedings of the 7th ACM SIGCOMM conference on Internet measurement - IMC '07, 2007Co-Authors: Phillipa Gill, Martin Arlitt, Anirban MahantiAbstract:This paper presents a Traffic Characterization study of the popular video sharing service, YouTube. Over a three month period we observed almost 25 million transactions between users on an edge network and YouTube, including more than 600,000 video downloads. We also monitored the globally popular videos over this period of time. In the paper we examine usage patterns, file properties, popularity and referencing characteristics, and transfer behaviors of YouTube, and compare them to traditional Web and media streaming workload characteristics. We conclude the paper with a discussion of the implications of the observed characteristics. For example, we find that as with the traditional Web, caching could improve the end user experience, reduce network bandwidth consumption, and reduce the load on YouTube's core server infrastructure. Unlike traditional Web caching, Web 2.0 provides additional meta-data that should be exploited to improve the effectiveness of strategies like caching.
Alexandr Vesselkov - One of the best experts on this subject based on the ideXlab platform.
-
WF-IoT - Cellular IoT Traffic Characterization and Evolution
2019 IEEE 5th World Forum on Internet of Things (WF-IoT), 2019Co-Authors: Benjamin Finley, Alexandr VesselkovAbstract:The adoption of Internet of Things (IoT) technologies is increasing and thus IoT is seemingly shifting from hype to reality. However, the actual use of IoT over significant timescales has not been empirically analyzed. In other words the reality remains unexplored. Furthermore, despite the variety of IoT verticals, the use of IoT across vertical industries has not been compared. This paper uses a two-year IoT dataset from a major Finnish mobile network operator to investigate different aspects of cellular IoT Traffic including temporal evolution and the use of IoT devices across industries. We present a variety of novel findings. For example, our results show that IoT Traffic volume per device increased three-fold over the last two years. Additionally, we illustrate diversity in IoT usage among different industries with orders of magnitude differences in Traffic volume and device mobility. Though we also note that the daily Traffic patterns of all devices can be clustered into only three patterns, differing mainly in the presence and timing of a peak hour. Finally, we illustrate that the share of LTE-enabled IoT devices has remained low at around 2% and 30% of IoT devices are still 2G only.
-
WF-IoT - Cellular IoT Traffic Characterization and Evolution
2019 IEEE 5th World Forum on Internet of Things (WF-IoT), 2019Co-Authors: Benjamin Finley, Alexandr VesselkovAbstract:The adoption of Internet of Things (IoT) technologies is increasing and thus IoT is seemingly shifting from hype to reality. However, the actual use of IoT over significant timescales has not been empirically analyzed. In other words the reality remains unexplored. Furthermore, despite the variety of IoT verticals, the use of IoT across vertical industries has not been compared. This paper uses a two-year IoT dataset from a major Finnish mobile network operator to investigate different aspects of cellular IoT Traffic including temporal evolution and the use of IoT devices across industries. We present a variety of novel findings. For example, our results show that IoT Traffic volume per device increased three-fold over the last two years. Additionally, we illustrate diversity in IoT usage among different industries with orders of magnitude differences in Traffic volume and device mobility. Though we also note that the daily Traffic patterns of all devices can be clustered into only three patterns, differing mainly in the presence and timing of a peak hour. Finally, we illustrate that the share of LTE-enabled IoT devices has remained low at around 2% and 30% of IoT devices are still 2G only.
