The Experts below are selected from a list of 71367 Experts worldwide ranked by ideXlab platform
Peter Steenkiste - One of the best experts on this subject based on the ideXlab platform.
-
an empirical analysis of a large scale mobile cloud Storage Service
Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
-
Internet Measurement Conference - An Empirical Analysis of a Large-scale Mobile Cloud Storage Service
Proceedings of the 2016 Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
Xiaohui Wang - One of the best experts on this subject based on the ideXlab platform.
-
an empirical analysis of a large scale mobile cloud Storage Service
Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
-
Internet Measurement Conference - An Empirical Analysis of a Large-scale Mobile Cloud Storage Service
Proceedings of the 2016 Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
Gaogang Xie - One of the best experts on this subject based on the ideXlab platform.
-
an empirical analysis of a large scale mobile cloud Storage Service
Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
-
Internet Measurement Conference - An Empirical Analysis of a Large-scale Mobile Cloud Storage Service
Proceedings of the 2016 Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
Ningjing Huang - One of the best experts on this subject based on the ideXlab platform.
-
an empirical analysis of a large scale mobile cloud Storage Service
Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
-
Internet Measurement Conference - An Empirical Analysis of a Large-scale Mobile Cloud Storage Service
Proceedings of the 2016 Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
Jianer Zhou - One of the best experts on this subject based on the ideXlab platform.
-
an empirical analysis of a large scale mobile cloud Storage Service
Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
-
Internet Measurement Conference - An Empirical Analysis of a Large-scale Mobile Cloud Storage Service
Proceedings of the 2016 Internet Measurement Conference, 2016Co-Authors: Xiaohui Wang, Ningjing Huang, Mohamed Ali Kaafar, Jianer Zhou, Gaogang Xie, Peter SteenkisteAbstract:Cloud Storage Services are serving a rapidly increasing number of mobile users. However, little is known about the differences between mobile and traditional cloud Storage Services at scale. In order to understand mobile user access behavior, we analyzed a dataset of 350 million HTTP request logs from a large-scale mobile cloud Storage Service. This paper presents our results and discusses the implications for system design and network performance. Our key observation is that the examined mobile cloud Storage Service is dominated by uploads, and the vast majority of users rarely retrieve their uploads during the one-week observation period. In other words, mobile users lean towards the usage of cloud Storage for backup. This suggests that delta encoding and chunk-level deduplication found in traditional cloud Storage Services can be reasonably omitted in mobile scenarios. We also observed that the long idle time between chunk transmissions by Android clients should be shortened since they cause significant performance degradation due to the restart of TCP slow-start. Other observations related to session characteristics, load distribution, user behavior and engagement, and network performance.
